Choshu-yuki banned from Seki games

[Shiyonofuji/Kaiopectate 2]

ISP is the best game run by the best people, period!! (Yusho winner...) (Showing respect...) (Clapping wildly...)

(In a state of confusion...) Medication?

I remember the distant past, when he was humorless and humble.

[Doitsuyama 1,192]

I close by saying, I embrace you all, as we commence 2011. But, right now, I shake my imaginary head. I'm going to go scratch my real dogs' heads for a while and let them lick my face all they want.

Kaiopectate, TK(but not really)SM

You must have been really shaken up as you submitted a hoshitori entry with the shikona Kaiopectate shortly after writing this and one day after an entry with Shiyonofuji... now I have deleted the Kaiopectate entry, but if you tell me this was intended as a changed entry for Shiyonofuji I will make the update.

[Kintamayama 45,367]

ISP is the best game run by the best people, period!! (Yusho winner...) (Showing respect...) (Clapping wildly...)

(In a state of confusion...) Medication?

I remember the distant past, when he was humorless and humble.

You forgot the third H- Healthy.

[Shiyonofuji/Kaiopectate 2]

I close by saying, I embrace you all, as we commence 2011. But, right now, I shake my imaginary head. I'm going to go scratch my real dogs' heads for a while and let them lick my face all they want.

Kaiopectate, TK(but not really)SM

You must have been really shaken up as you submitted a hoshitori entry with the shikona Kaiopectate shortly after writing this and one day after an entry with Shiyonofuji... now I have deleted the Kaiopectate entry, but if you tell me this was intended as a changed entry for Shiyonofuji I will make the update.

The latter was intended to correct the former, so please do, if it's not too much of a hassle. But don't worry about it unduly either way. It's not as though the latter is some clever improvement or conniving doppelganger ploy.

I have never registered for or competed in hoshitori as anything other than Shiyonofuji. My hoshitori history as Shiyonofuji, which extends for 53 basho back to 1997, excluding gaps, may be seen at http://sumodb.sumogames.com/hoshi/HoshPlayer.aspx?p=240

It is a relatively inept experience in which I take great pride. I enjoy entering and highly respect the predictive abilities of the consistent leaders. In contrast, there is no hoshitori listing at all for Kaiopectate, who only plays Bench -- which Shiyonofuji does not. The legend of the noisy birth of Kaiopectate, after Shiyonofuji was barred from late entry to a Bench basho due to some absurd hidden-in-plain-sight midweek deadline which he forgot, is well known to me.

I know Kaiopectate. I know Shiyonofuji. Neither, sir, is a Norizo. Nor a Chosho-yuki, for that matter. (Interestingly, "Chosho-yuki," in English, is translated as "chosho-yuki"!

I don't think you're suggesting that I intentionally (and surreptitiously) attempted to submit two entries -- after all, I used the same e-mail address. Or to conceal the dual identities: see my dual-shikona registration on this site. I openly confess to being an oblivious idiot but not to cheating. However, if you feel that my error represents a breach of some kind, please just let me know.

[Kintamayama 45,367]

I don't think you're suggesting that I intentionally (and surreptitiously) attempted to submit two entries -- after all, I used the same e-mail address. Or to conceal the dual identities: see my dual-shikona registration on this site. I openly confess to being an oblivious idiot but not to cheating. However, if you feel that my error represents a breach of some kind, please just let me know.

See you at sundown. I'll bring the blindfold.

[Doitsuyama 1,192]

The latter was intended to correct the former, so please do, if it's not too much of a hassle.

I updated your hoshitori entry now.

[Shiyonofuji/Kaiopectate 2]

The latter was intended to correct the former, so please do, if it's not too much of a hassle.

I updated your hoshitori entry now.

Vielen dank, on behalf of both!

[Flohru 176]

Thinking the Norizo-CY-case over for some time I would like to add my thoughts.

While I'm convinved that Asashosakari did a thorough statistical analysis, I think there is simply no way to prove "guilt" or "innocence" with stats in cases like that. Just two examples:

C-Y's selections across multiple games are wildly inconsistent.

This seems to me not suspicious at all - many players are using this strategy, you just need to have a look at Doitsuyama's ISP picks in the last couple of years...

And on that topic: While running drone accounts for "fun" probably isn't a completely unusual occurence, I get highly annoyed when they're used to enable a player multiple shots at the yusho, and here we find my personal last straw: With C-Y in the Quad yusho race in Kyushu after the first few days, the similarity between the N and C-Y entries skyrocketed beyond its already high long-term level: While Day 1-5 saw 55% identical picks (11 of 20), which is about the 2009-2010 average, the number went to 69% for Day 6-9 (11 of 16), and unbelievable 85% for Day 10-14 (17 of 20). This indicates that, with a yusho at stake for the drone, Norizo felt compelled to copy his own "first-choice" picks onto the drone account almost verbatim, as it was clearly no time for experiments.

This interpretation is just Asashosakari's own assumption and there are of course other possible explanations for the fact that N and CY picks were becoming more similar towards the end of the mentioned basho in SQ, the most obvious being that CY herself decided that it is time to change her strategy in order to win the yusho, thus picking more conservative (and more like N) in the final days...

Also, you could probably find suspicious stats like that for all "family players" discussing the bouts together (as indicated by Umigame). Sometimes their picks will be similar, sometimes not, I could even imagine that in those cases picks tend to be more similar towards the end of the basho, when the performances of the rikishi can be discussed more substantially.

It seems to me that the only difference is that N and CY are fairly successful, so their case catches one's eye. Banning a long-time sumo games player like CY without even consulting her or N and making this all public without real evidence seems to me highly inappropriate, despite dead email-addresses and everything - what about a PM or a message to some other Japanese player who is in contact with N/CY (Susanoo?)...

And finally: I think it's preposterous for other game admins to simply jump on the bandwagon without looking into the N/CY-case more thoroughly in their own game. Is there any evidence that CY and N picks e.g. in UDH or RotoSumo are made by only one person? And how would you prove something like that in those games anyway? If the picks are similar, maybe they are simply discussing their picks prior to the basho like many "family players" are doing for sure, and if they are not, could this not merely indicate that CY and N are indeed two different players/persons with their own preferences and game strategies?

Edit: And one more point I forgot:

Take the Oracle picks of Haru 2009. In Juryo, there had been 27 rikishi in this basho. In 25 cases, the picks of Norizo and Choshu-yuki have been identical (with 1 win difference in the other two picks). I find it hard to believe that this similarity just comes from talking about the rikishi.

This seems indeed intriguing the way you present it, but I think you should mention that both of them were using the "pick only 7/8 wins for all juryo rikishi except one or two"-strategy in that basho, a strategy I've also used a few times - so in fact the only suspicious thing beside both of them trying the same strategy (which rather indicates that it is not one player entering two basically identical entries) is that they had in 26 out of 27 cases the same KK/MK prediction - but then Norizo and I shared 24 out of 28 in Oracle last basho without ever discussing Sumo...

Edited January 10, 2011 by Flohru

[Randomitsuki 2,831]

Thinking the Norizo-CY-case over for some time I would like to add my thoughts.

Thank you for your input.

Also, you could probably find suspicious stats like that for all "family players" discussing the bouts together (as indicated by Umigame). Sometimes their picks will be similar, sometimes not, I could even imagine that in those cases picks tend to be more similar towards the end of the basho, when the performances of the rikishi can be discussed more substantially.

As I said before, we are not talking about mere similarity. To illustrate the point, here are Quad data from the last two years.

On the x-axis you see the percentage of similarity between player pairs, and on the y-axis you see the number of pairs for which this percentage was found. All in all you get a regular bell shape. The tiny dot that you see at 55% similarity, is the Norizo-Choshu-yuki pair. It's more than five standard deviations above the mean. The probability for this to happen by chance is about 1 against 14 million. And this is just for one of the games.

It's bad enough that three games were analyzed, each with thousands of potential player pairs, and in all three games they had the most similar picks - but the degree of similarity is the really flabbergasting thing.

It seems to me that the only difference is that N and CY are fairly successful, so their case catches one's eye.

But the analyses for the three games in question was conducted for all possible player pairs. And while sometimes (well, even usually) family members or close friends have an above average similarity, all these similarities pale in comparison to that of Norizo and Choshu-yuki.

Later edit: One additional comment about the Quad example from above. There are some players who really have an eye for "average" picks. For instance, Meyeryu is "responsible" for nine player similarities at or above 40%. His picks are hitting public taste, and therefore he has a high average of similarity with many other players. However, this is neither the case for Norizo or Choshu-yuki. The second-most similar player to Norizo is at 32%, the second-most similar player to Choshu-yuki ranges at 30%, well within the bell curve.

Edited January 10, 2011 by Randomitsuki

[Flohru 176]

Thanks, Randomitsuki, for further illustrating the strong statistical foundation of your accusation. I see that you even found a convincing way to show that the degree of similarity between CY and N in Quad is without comparison so my argument here is refuted.

Still as initially mentioned I think those stats are intriguing but they are no definite proof for something fishy going on. Heck, you could even argue that CY knowing N to be the most successful sumo games player ever is more likely to listen closely to what her husband (?) is telling her at the evening table regarding the matches of the next day or the rikishi in general than any one else in those family playing cases...

[Asashosakari 19,506]

FWIW, the most charitable interpretation I'm inclined to believe without further explanations would be this:

Maybe Norizo and Choshu-yuki enter their picks separately, but I am convinced that one of them has access to the picks of the other. Take the Oracle picks of Haru 2009. In Juryo, there had been 27 rikishi in this basho. In 25 cases, the picks of Norizo and Choshu-yuki have been identical (with 1 win difference in the other two picks). I find it hard to believe that this similarity just comes from talking about the rikishi. Maybe Norizo writes the picks down on paper or has them in a computer file, and maybe Choshu-yuki copies his picks and makes very few changes (or vice versa). In any case, this is not independent picking anymore. It's copying plus minor tweaking.

I'm open to the idea that Choshu-yuki's Toto picks don't actually reflect her true opinions (which we are told are practically the same as Norizo's), which would presumably imply that she's playing strictly for the kanto-sho. Can't say I'm enamoured with that possibility, but nothing shady about that, so let's take it as given. Even so, she's not simply picking the upset in every bout, e.g. she's usually not betting against Hakuho, so one has to assume that upsets are only picked when she believes there is actually a credible chance of them happening.

The most astounding thing happened in S4 last basho on Day 15. I'm too lazy to generate a table, so a screencap from my spreadsheet will have to suffice:

At this point each had 12 rikishi remaining, 10 of them identical (everybody not marked "X" in the S4 columns). As you can see Norizo's Toto picks projected 2 winners for him, Yoshikaze and Bushuyama. In addition he selected Homasho and Kotokasuga for S4. Since we're operating under the assumption that Choshu-yuki's real opinions match Norizo's, we compare her S4 slate to Norizo's Toto picks, not her own, and find three winning rikishi: Yoshikaze again, plus Kaisei and Sakaizawa. Sure enough, these three rikishi show up in her S4 picks, even though her own Toto picks project all three to be losing. That's okay, we're assuming she simply picked their opponents for potential upsets, and she actually believes that Yoshikaze and (especially) Kaisei and Sakaizawa should be winning.

However, her S4 lineup still needs a fourth rikishi now. It turns out to be - Kasugao. How odd, that's a rikishi she wasn't even willing to predict as an upset winner! Gone unpicked are no less than 7 rikishi she did feel capable of an upset. At this point we're left with two possible explanations: 1) Choshu-yuki's Toto picks are deeply irrational; she did think Kasugao was a good upset pick, but she didn't select him as such. Or 2) She deferred to Norizo's opinion even in a situation where her own opinion provided additional, crucial information, namely which of those many losers from Norizo's Toto picks she felt were most likely to cause an upset. (Keep in mind that she was in the S4 yusho race at this point.) Since I don't believe in ascribing irrationality to people when it comes to simple choices, I'm forced to go with 2. But if that's the case, I'm not willing to consider Choshu-yuki an independent player; her picks are then mere reflections of Norizo's thoughts.

Okay, and to be fair, perhaps there's a third possibility: She doesn't actually have any type of consistent thinking from one game to the next, and Kasugao was picked completely at random. But that would only reinforce the notion that her selections are simply slight reworkings of Norizo's, as Randomitsuki surmised in the section I quoted above. Some people have called her "a good player" in this thread; I assume that assessment would crumble under that scenario. And that's the very issue here - at this point, nobody knows who Choshu-yuki actually is, in terms of game performances.

(In case you're wondering where the heck that Kasugao pick came from if it wasn't random, considering that Norizo himself selected Homasho and Kotokasuga from among his available losers and C-Y did have Kotokasuga available to mirror this, I have a theory: Kasugao and Bushuyama faced each other. As you can see, Bushuyama - projected as a winner - was not available to Choshu-yuki. However, selecting his opponent means that, should Norizo be wrong about Bushuyama's winning prospects, at least Choshu-yuki automatically grabs that point. If true, this is again a lot more consistent with the notion of Norizo calling all shots than it is with Choshu-yuki having at least some mind of her own when it comes to her S4 picks.)

Edit: And again, I'm very interested in finding out how a player whose recent Toto attendance looks like this can have an S4 scoresheet that looks like this. The difference in missed days is a little too striking, in my opinion.

Edited January 10, 2011 by Asashosakari

[Asashosakari 19,506]

For completeness:

It seems to me that the only difference is that N and CY are fairly successful, so their case catches one's eye. Banning a long-time sumo games player like CY without even consulting her or N and making this all public without real evidence seems to me highly inappropriate, despite dead email-addresses and everything - what about a PM or a message to some other Japanese player who is in contact with N/CY (Susanoo?)...

Tried, and failed. I don't want to speak for the relevant party, but the explanation I was given recently points to an unfortunate case of bad timing, and perhaps things would have worked out differently if everything had happened three months earlier or three months later. Still, we did make considerable efforts to discover a way of contacting Norizo for the better part of two months.

BTW, if the analysis carried out on S4 and Oracle had - by accident, as it were - discovered other pairs of players with completely implausible similarities, they would have been removed from the games, too. Yes, family members do show increased levels of selection similarities in that analysis, but nothing remotely close to N/C-Y.

Also, I'm completely open to re-instating Choshu-yuki in the games, but as Randomitsuki said up-thread, this will require major explanations of what the heck is actually going on to create these implausible patterns.

For visual edification, the top 10 S4 similarity percentages (without names)...data from 2009.01-2010.09, minimum 200 mutual picks played (out of a maximum of 660):

55.3%

43.9%

42.6%

42.1%

41.9%

41.6%

41.1%

40.8%

40.7%

40.6%

I defy people to not question such a major outlier.

Edited January 10, 2011 by Asashosakari

[Flohru 176]

I defy people to not question such a major outlier.

I think nobody did question your statistical points, but one may very well question the conclusions you draw from these numbers.

Moreover as they seem to be not really clear-cut themselves - in your initial post you basically told us that the data indicates that CY is simply a drone account of N which led e.g. Zentoryu to ban CY from RotoSumo "because duplicate accounts are not allowed."

But now it sounds as if you are following Randomitsuki's (IMHO far more likely) explanation that CY only has access to the picks of N. That's a major difference as I have yet to see a rule in any sumo game requiring "independent picking" of the players, whatever that may be. Some players may listen to the advice others give them or the information they gain from the internet or their connections to the real Sumo world and others may not, some players will take a look at how other players pick e.g. in DST or ISP and some may not - who is the person to judge if the degree to which they are e.g. listening to what their partner is telling them at dinner is still legitimate or not. Heck, I think we even have infants participating in some of the sumo games, how could we except them to enter their picks independently??

So as CY really seems to be the existing individual we all believed her to be and she simply benefits from knowing in one way or the other what N is doing, according to the rules there is no way to ban her even if the picks of CY and N are way more similar than those of any other two players. And then again I assume this only holds true for some of the games anyway and CY should definitely not be banned in e.g. RotoSumo, unless you find some examples for identical or semi-identical entries here. And even if you did there are always cases of identical picks by completely unaffiliated players in games like RotoSumo...

To sum up, I don't think it's N duty to explain the stats (in the first place wouldn't that be CY duty anyway?), but yours to prove without any doubt that N violated the existing rules by using multiple entries which is the only case that allows for such a drastic sanction as banning a long-time player.

[Doitsuyama 1,192]

And then again I assume this only holds true for some of the games anyway and CY should definitely not be banned in e.g. RotoSumo, unless you find some examples for identical or semi-identical entries here. And even if you did there are always cases of identical picks by completely unaffiliated players in games like RotoSumo...

True. I did some statistical analysis for the hoshitori game of course, and the same outcome here, Norizo's and Choshu-Yuki's picks were too likely to be a product of (even loosely) independent players.

To sum up, I don't think it's N duty to explain the stats (in the first place wouldn't that be CY duty anyway?), but yours to prove without any doubt that N violated the existing rules by using multiple entries which is the only case that allows for such a drastic sanction as banning a long-time player.

Hmm, CY seems to be a somewhat mythical figure in that sense, especially for such a knowledgable person. I have never read anything from her, but I have read from Norizo here and earlier in Japanese forums. Anyone please correct me here if I am mistaken. Look, if CY would be a distinctive member of the sumo gaming community, I'm sure there would be much more leeway from me, but it still would require an explanation for the similar entries.

P.S.: My (rather rare) deviations in ISP and Toto picks are clearly strategic as they only happen when I pick the underdog in ISP. (Clapping wildly...)

[Randomitsuki 2,831]

Prediction:

pointless discussions and a protracted quasi-legal battle about just how strong the statistical evidence is

Outcome:

So as CY really seems to be the existing individual we all believed her to be and she simply benefits from knowing in one way or the other what N is doing, according to the rules there is no way to ban her even if the picks of CY and N are way more similar than those of any other two players.

The reason why multiple entries for one player are not allowed is because these picks are not independent from each other. If Norizo and Choshu-yuki discuss all bouts and all pre-basho picks together, there is such a strong dependence that I would very much prefer if they prepared just one "Norizo+Choshu-yuki" pick. That would be fair, and I hope that's the way these two (if they are both still interested in playing) will handle that issue.

In fact, if I made my entries using my spreadsheets, and if someone took away these spreadsheets and asked me to re-enter the same picks from memory a few hours later, there would be a high probability that my two picks are less similar to each other than the picks of two players who go through all bouts together, and then settle minor disagreements via slightly different picks.

Collaboratively going through all sumo gaming decisions, and then making two entries out of it (probably even having access to the picks by the other one), is something that I would like to forbid if it happened on a systematic basis.

Of course, it is difficult to draw a line between "talking about every bout", "talking about some bouts", and the like. But the statistical evidence in our particular case points at a collaboration that is so closely knit that I find it highly unfair to make two different entries out of it.

Edited January 11, 2011 by Randomitsuki

[Flohru 176]

Prediction:

pointless discussions and a protracted quasi-legal battle about just how strong the statistical evidence is

Outcome:

So as CY really seems to be the existing individual we all believed her to be and she simply benefits from knowing in one way or the other what N is doing, according to the rules there is no way to ban her even if the picks of CY and N are way more similar than those of any other two players.

Sorry, but I can't see any pointless discussions going on - you've banned a long-time player, something which never happened before I think, because of strong statistical evidence but you could very well expect some others to question the conclusions you have drawn from those stats and the harsh decision itself which only seems to be justified by your own interpretation of the "multiple entries are forbidden"-rule (personally I don't think it's necessary at all for players who are discussing the bouts together and agreeing most of the time to prepare just one entry, thus possibly quarreling a lot about existing differences). Morevoer as this mess could easily lead to Norizo retiring which would be a major blow for all games, and it could set a precedent, as Asashosakari already indicated, in case you will ever detect "suspicious" picking patterns again...

Well, I've stated my case so I'll be quiet now.

[Randomitsuki 2,831]

Prediction:

pointless discussions and a protracted quasi-legal battle about just how strong the statistical evidence is

Outcome:

So as CY really seems to be the existing individual we all believed her to be and she simply benefits from knowing in one way or the other what N is doing, according to the rules there is no way to ban her even if the picks of CY and N are way more similar than those of any other two players.

Sorry, but I can't see any pointless discussions going on

I was actually referring to the words "quasi-legal battle" rather than to the word "pointless". Sorry for that!

[Asashosakari 19,506]

Sorry, didn't have time to get back to the thread until now.

Yes, you're free to quibble with my use of the words "drone" and "sock puppet", but if one player essentially dictates a second set of picks to another person and then has that person enter these selections, it really makes no difference to me if this arrangement happens to be entirely voluntary. And I strongly disagree that Norizo has no explaining to do, given the extraordinary nature of the picks that have resulted from whatever it is that is going on. (I'll note that in that Quad top 10 I posted, all those 40%ers include no other known family connections.) In any case, this isn't a court of law, and playing Seki-[Toto/Quad/Oracle] is not a right. I'm certainly not going to paralyze myself in my best-effort enforcement of fair play by adhering to a standard of "without a doubt"; you know as well as I do that that's impossible without a direct confession. As I said before, I'm completely ready to reinstate Choshu-yuki to her previous ranks* if a convincing explanation comes forward. But explained it must be.

* Which, incidentally, would even include a repromotion to ozeki in Oracle, joining Norizo again, thanks to her yusho-doten in Kyushu. It's almost a shame that she failed to win the yusho tiebreak against you, Flohru, as I'd have been extremely curious about your reaction if all these suspicions turn out to be justified and she had taken away a yusho from you in her final appearance. :-P

Further random comments:

The reason why multiple entries for one player are not allowed is because these picks are not independent from each other. If Norizo and Choshu-yuki discuss all bouts and all pre-basho picks together, there is such a strong dependence that I would very much prefer if they prepared just one "Norizo+Choshu-yuki" pick. That would be fair, and I hope that's the way these two (if they are both still interested in playing) will handle that issue.

I almost added a similar comment to my previous post (#36), but felt it might be a bit too flippant. Yes, if Norizo and Choshu-yuki (or really, any number of players greater than 1) want to play "consensus picks", they're free to make it official and play under a single account.

Hmm, CY seems to be a somewhat mythical figure in that sense, especially for such a knowledgable person. I have never read anything from her, but I have read from Norizo here and earlier in Japanese forums. Anyone please correct me here if I am mistaken. Look, if CY would be a distinctive member of the sumo gaming community, I'm sure there would be much more leeway from me, but it still would require an explanation for the similar entries.

Yes, that does play a strong role in my "nobody knows who Choshu-yuki actually is, in terms of game performances" comment. Is she the daring risk-taker seen in Toto? Or the conservative player seen in Oracle and Quad? How come even games with fundamentally similar mechanics (such as Hoshitori + Roto) exhibit diverging patterns, at least as far as my anecdotal observations go? (We didn't run any detailed analysis on Roto.) The kanto-sho in Toto certainly provides incentive to go risky, but what's the incentive in Rotosumo? Choshu-yuki is stuck in makushita there - in Hoshitori she's gone to yokozuna by mimicking (if that's what it is) Norizo's entries. It's pretty clear which strategy is the superior one.

My suspicion remains that Choshu-yuki does play in some games still, and perhaps did play in some others that are strongly under suspicion now but has since passed on most/all of the decision-making for those. If C-Y is genuine in Toto, but not in Quad/Oracle, I'll reinstate her in Toto, no problem. But until some type of explanation to that effect comes forward, I reserve to treat all her game entries as equally suspicious. In that vein: I'm not concerned what other game admins decide to do with the information posted in this thread as that's obviously their decision and theirs alone, but I for one can't fault those who seem to share my "it's all suspect now" mindset and have gone ahead with their own temporary removal measures for Hatsu basho. YMMV, of course.

And even if you did there are always cases of identical picks by completely unaffiliated players in games like RotoSumo...

Well, hopefully not the same players for 11 basho in a row, or I'd get suspicious even in a game with somewhat limited decision space like Roto.

Edited January 12, 2011 by Asashosakari

[Flohru 176]

Well okay, two more small comments:

* Which, incidentally, would even include a repromotion to ozeki in Oracle, joining Norizo again, thanks to her yusho-doten in Kyushu. It's almost a shame that she failed to win the yusho tiebreak against you, Flohru, as I'd have been extremely curious about your reaction if all these suspicions turn out to be justified and she had taken away a yusho from you in her final appearance. ;-)

Believe it or not, but the time I would have cared about this is long gone. Some years ago when I was about to retire because of (a perceived and quite frustrating) lack of success, I decided to just play for fun from now on which was a wise decision and is maybe the main reason why I'm so unhappy about seeing other players banned...

My suspicion remains that Choshu-yuki does play in some games still, and perhaps did play in some others that are strongly under suspicion now but has since passed on most/all of the decision-making for those.

If that is your suspicion, don't you consider it problematic that following your initial post indicating her to be a mere "sock puppet account" of N, CY got banned by other game admins as well though nobody (except for Doitsuyama backing up your decision and Gusoyama who subsequently decided not to ban CY for now) ever really investigated if there is anything suspicious about her "picking behaviour" in e.g. UDH or RotoSumo?

Edit: Doitsuyama is of course somebody and should be mentioned...

Edited January 13, 2011 by Flohru

[Doitsuyama 1,192]

got banned by other game admins as well though nobody (...) ever really investigated if there is anything suspicious about her "picking behaviour"

Am I nobody? Oh, ok... ;-)

[Asashosakari 19,506]

Believe it or not, but the time I would have cared about this is long gone. Some years ago when I was about to retire because of (a perceived and quite frustrating) lack of success, I decided to just play for fun from now on which was a wise decision and is maybe the main reason why I'm so unhappy about seeing other players banned...

Good for you that your enjoyment of the games can't be marred by unfair behaviour of other players, I guess... Maybe all game admins should just say, "screw it, anything goes from now on!" and we'll have the most enjoyable games ever. We could award sansho for the most clever cheating! ;-) Sorry to be sarcastic, but I really find it difficult to understand how it's possible to be so undisturbed. I would guess you'd feel differently if you were running a game yourself.

If that is your suspicion, don't you consider it problematic that following your initial post indicating her to be a mere "sock puppet account" of N, CY got banned by other game admins ...

Quite honestly, I lost most of my concerns about such potential side effects after spending two unsuccessful months trying to find a way to solve this issue in a non-public way. Maybe I'm a jerk, but when contacting the email address on file at various games results in this:

Hi. This is the qmail-send program at [my mail server].

I'm afraid I wasn't able to deliver your message to the following addresses.

This is a permanent error; I've given up. Sorry it didn't work out.

<[address redacted]@ybb.ne.jp>:

203.216.243.174_failed_after_I_sent_the_message./Remote_host_said:_554_delivery_error:_dd_This_user_doesn't_have_a_ybb.ne.jp_account_([address redacted]@ybb.ne.jp)_[-100]_-_ybbmta84.mail.tnz.yahoo.co.jp/

and contacting the email address connected to his own game results in this:

Hi. This is the qmail-send program at [my mail server].

I'm afraid I wasn't able to deliver your message to the following addresses.

This is a permanent error; I've given up. Sorry it didn't work out.

<sumogames@[major portal site].jp>:

203.190.60.202_does_not_like_recipient./Remote_host_said:_550_5.7.1_<sumogames@[major portal site].jp>:_Recipient_address_rejected:_Unknown_User/Giving_up_on_203.190.60.202./

forcing the issue to be made public or not addressed at all, I can't say I'm overly concerned about potential overreach by third parties. At some point you just have it coming. :-S And as I said, speaking simply for myself I consider all C-Y game entries to be under suspicion now until evidence to the contrary is presented: If I was somehow in the position of running all existing sumo games, she would have been banned from all of them even if statistical evidence had only been ascertained for a subset of them. Finding irregularities in one game may point to only a singular case - finding irregularities in 3+ calls everything else into question, as well. Again, YMMV.

At the same time, I trust that if a convincing explanation comes forward, all concerned admins (assuming they, too, find it convincing) will reinstate Choshu-yuki as will I do. At this point I'm the only really exposed admin anyway as I've already removed C-Y from the banzuke*; for all other games it's basically an enforced kosho.

* Okay, and Hoshitori, but that banzuke doesn't have any issues with scarcity of ranks.

Maybe I've made a mistake by targetting only Choshu-yuki for sanction, under the assumption that it's a "lesser" account than Norizo. It remains to be seen if perhaps I was too charitable in that interpretation and both accounts should be considered equally tainted instead.

Edited January 13, 2011 by Asashosakari

[yorikiried by fate 2,029]

OK, my doctor said I shouldn't get involved, but seeing four fellow Germans taking things lightly, as we are wont to do, I just can't keep it shut.

Hausrecht is the key word here. These games are not exercises in democracy. Somebody runs them. He/she's the tyrant. It is a huge act of mercy to not just shut off both accounts without even explaining the how and why. I would have done it for the dead email adresses reason alone.

[Flohru 176]

Good for you that your enjoyment of the games can't be marred by unfair behaviour of other players, I guess... Maybe all game admins should just say, "screw it, anything goes from now on!" and we'll have the most enjoyable games ever. We could award sansho for the most clever cheating! ;-) Sorry to be sarcastic, but I really find it difficult to understand how it's possible to be so undisturbed. I would guess you'd feel differently if you were running a game yourself.

Was it just your desire to be sarcastic or is it really that difficult to understand me?

Nobody likes unfair behaviour of other players and if you can prove that N is using the CY account in order to make duplicate entries then, according to the rules, please do ban all involved accounts.

If you can just prove that CY knows in some way about N picks and is influenced by that knowledge, thus making their picks (very) similar to each other, then I'm not fussed about this at all, and I think a majority of players would agree with me here, though maybe not the ones likely to read this thread. But yes, it's not a democracy and in the end it's up to the game admins to do whatever they like.

Sumo games (as all such games) were made to increase interest in the sport and - where possible - encourage discussions about the rikishi and their performance, thus I can't see anything wrong if family members or friends are entering similar picks, provided that they are real persons interested in participating and not just fake accounts, of course.

[Kintamayama 45,367]

I first wish to state that I don't (as usual) understand all the graphs and stats, but I am sure a lot of time and effort went into the investigation and everything was done honestly etc.. etc..

I do know this-for many years, many players entered their children's guesses in Bench Sumo (and a list was made so they won't fight each other, as heyamates..). Sometimes the children were very young, by their own admission. I of course can't prove that the children didn't actually enter the picks, but well.. IIRC, just recently we had some new players who were offspring of current players. As for similarity in picks, I think it's the OPPOSITE- if I were playing two shikona, I would play them differently-otherwise what's the point?

I have to go now, I'm busy.

[Asashosakari 19,506]

I do know this-for many years, many players entered their children's guesses in Bench Sumo (and a list was made so they won't fight each other, as heyamates..). Sometimes the children were very young, by their own admission. I of course can't prove that the children didn't actually enter the picks, but well..

That's quite a different issue though, isn't it? The point isn't who's entering the picks, but who's selecting them. I'm always skeptical in such cases about the degree to which the child is actually responsible for the selections, but... (The "my one-year-old child is playing with dice-rolled selections" case seen among the current crop of Bench Sumo rookies is kinda borderline for me...I think I'd balk if somebody wanted to do that for two children.) Anyway, what Randomitsuki and I are harping on about here is one mind potentially creating the selections for two "players". But I for one don't care one bit about who's actually sending the selections...if an entire soccer lineup want to choose one point-of-contact guy for entering their Quad picks, they're free to do so, as long as each of those 11 players used their own brain to come up with their choices.

As for similarity in picks, I think it's the OPPOSITE- if I were playing two shikona, I would play them differently-otherwise what's the point?

Well, if I was playing with two shikona just for the giggle factor, that's what I would do, too. If I'm an above-average player and actually want to be successful with both accounts, the entries would probably be quite a bit more similar. Like, say, sharing at least 2 out of 4 Quad picks every day...