Invitation to Play GTB - March 2025 - 359 entries - RESULTS!!

[Andonishiki 170]

let's assume we know the 42 guys....

should the mathematically number to take all possibilities into account be

42x41x40x39x38x.......x3x2 =
4.21502E +51

probably NO WAY the software can process so many bets or any result list could show this number of participants....

but just in case - should there be a limit of new players for each basho (maybe 100 max) ?

 

[Reonito 1,469]

30 minutes ago, Sumo Spiffy said:

Mathematically it would be at least 42!, since technically anyone could go anywhere. That's assuming we can determine exactly which 42 guys will be in the division, which isn't the case this time (Shirokuma vs. Tamashoho). If we remove the single yokozuna from that equation, since his position is truly locked, it's 41!. Assume the ozeki as well, and it's 39!.

This is not a useful calculation in for thinking about GTB in practice though. For anyone of reasonable skill on a typical banzuke (excluding ones where sanyaku number is in doubt), there are about 15-20 slots that are completely nailed down. For the remaining (say) 25 there are usually only a couple of choices, and they're not independent (if guy A goes to 4W, guy B goes to 5E, and vice versa). Thinking about it as the odds of getting (say) 12 coin flips right (which is 1 in 4096) probably gets a lot closer to reality. Then throw in one or two official choices each time that deviate from all logic and are therefore much harder to guess than 50:50. The game has had something like 150 editions, with say 100 players on average, so that's 15,000 guesses. We've had scores as high as 78 (I think that's the top), one or two flips away from perfection, which makes sense under this very crude calculation but not if anyone could really go anywhere.

[Ryoshishokunin 301]

Yeah, the hard part is winning all the coinflips while being off-the-wall enough to catch the one or two unusual choices: it's unlikely that one player will do both. 

[Sumo Spiffy 568]

37 minutes ago, Reonito said:

This is not a useful calculation in for thinking about GTB in practice though. For anyone of reasonable skill on a typical banzuke (excluding ones where sanyaku number is in doubt), there are about 15-20 slots that are completely nailed down. For the remaining (say) 25 there are usually only a couple of choices, and they're not independent (if guy A goes to 4W, guy B goes to 5E, and vice versa). Thinking about it as the odds of getting (say) 12 coin flips right (which is 1 in 4096) probably gets a lot closer to reality. Then throw in one or two official choices each time that deviate from all logic and are therefore much harder to guess than 50:50. The game has had something like 150 editions, with say 100 players on average, so that's 15,000 guesses. We've had scores as high as 78 (I think that's the top), one or two flips away from perfection, which makes sense under this very crude calculation but not if anyone could really go anywhere.

Oh, for sure. The quoted message said "mathematically", so I went with the technical answer. It's also a lot easier to describe a number of possibilities in terms of factorials; I started to talk about limitations like no promotions for losing records, but then I was like... nah, lazy.

[Gurowake 4,084]

The main issue with anyone having a perfect match to the Kyokai's banzuke is that there is invariably something that makes us collectively scratch our heads and wonder why they decided that way.  While there tend to be a few players who might get that right, they'll generally have plenty of differences elsewhere because their logic is probably unsound in general to have come up with the weirdness.  I'm inclined to believe that they're intentionally making some weird choices just as a way of showing who's boss.

Edited February 17 by Gurowake

[Sumo Spiffy 568]

2 minutes ago, Gurowake said:

The main issue with anyone having a perfect match to the Kyokai's banzuke is that there is invariably something that makes us collectively scratch our heads and wonder why they decided that way.  While there tend to be a few players who might get that right, they'll generally have plenty of differences elsewhere because their logic is probably unsound in general to have come up with the weirdness.  I'm inclined to believe that they're intentionally making some weird choices just as a way of showing who's boss.

The first 84 will come from someone who says, "Oh man, I totally misclicked one of my picks!"

[Gurowake 4,084]

10 hours ago, Jakusotsu said:

Ok, I've checked: all 310 entries with different email accounts, and only two from the same IP address - presumably family.

All the cheaters are using VPNs obviously.

[Gurowake 4,084]

1 minute ago, Sumo Spiffy said:

The first 84 will come from someone who says, "Oh man, I totally misclicked one of my picks!"

Note you don't need 84 to get a perfect score in rare cases, like Kyushu 2021.

[Reonito 1,469]

1 hour ago, Sumo Spiffy said:

Oh, for sure. The quoted message said "mathematically", so I went with the technical answer. It's also a lot easier to describe a number of possibilities in terms of factorials; I started to talk about limitations like no promotions for losing records, but then I was like... nah, lazy.

If you want a more realistic but still very conservative upper bound for the current GTB, the top 4 ranks are locked. That leaves 38 slots to fill. I don't think there are more than 4 remotely plausible candidates for any of them, so assuming these are independent (which they aren't), that's 4^38 = 2^76 ~ 10^23. That the committee might put someone from outside the plausible 4 in a couple of slots is more than made up for by the fact that most slots have only one or two plausible choices. I'm not sure there were any where I seriously considered more than 3 options...

[Jakusotsu 5,981]

4 hours ago, Kintamayama said:

12 hours ago, Jakusotsu said:

Ok, I've checked: all 310 entries with different email accounts, and only two from the same IP address - presumably family.

That makes me officially a jerk, and we can happily move on.

I have 297 entries.. And how do you have access to all these details? I don't..

Right - sorry - although I was pruning the data for multiple entries by the same players (i.e. changes of mind) before comparing, I mistakenly counted all the entries while posting that number. Getting old...

[Asashosakari 19,394]

Just out of curiosity, how many entries have Terunofuji in them? 

[Sumo Spiffy 568]

2 hours ago, Gurowake said:

All the cheaters are using VPNs obviously.

I didn't jump into the conversation about this before, but I'mma be honest—if someone cares enough about putting in multiple entries to Guess the Banzuke that they're using a VPN to bypass our ability to detect them, so be it. A lot of people who see the "one entry per email" rule wouldn't think of using a VPN; they would just use a separate email. The fact we're not seeing any of that simpler method of trying to get around the rule means it's unlikely there's some broad, more technically-sound cheating issue either. Plus, the numbers go up when there's a push to get more people to play, and go down when there's not, which also indicates basically everyone who plays is playing as intended.

What strikes me as far more likely is one knucklehead with more time than sense using a VPN to blast off, like, twenty slightly different entries. Would that even create a problem? I don't know. Regardless, nothing suggests there's some group of a problematic size using VPNs to get around GTB rules. 

[Reonito 1,469]

12 hours ago, Gurowake said:

The main issue with anyone having a perfect match to the Kyokai's banzuke is that there is invariably something that makes us collectively scratch our heads and wonder why they decided that way.  While there tend to be a few players who might get that right, they'll generally have plenty of differences elsewhere because their logic is probably unsound in general to have come up with the weirdness.  I'm inclined to believe that they're intentionally making some weird choices just as a way of showing who's boss.

Edited February 18 by Jakusotsu

[Jakusotsu 5,981]

10 hours ago, Asashosakari said:

Just out of curiosity, how many entries have Terunofuji in them? 

Only one player with Terunofuji as Yw. Being greyed out on the sumodb banzuke apparently helped.

On the other hand there are four players with no yokozuna at all, so the news about Hoshoryu was easier to miss.

Weirdest entry is one player's with Hoshoryu as Yw and no Ye.

[Akinomaki 40,593]

11 hours ago, Sumo Spiffy said:

What strikes me as far more likely is one knucklehead with more time than sense using a VPN to blast off, like, twenty slightly different entries.

The only good reason to enter as multiple players is to get the number of players to new heights, as a present to Kintamayama

[Kintamayama 45,158]

48 minutes ago, Akinomaki said:

The only good reason to enter as multiple players is to get the number of players to new heights, as a present to Kintamayama

Yes indeed, but there forever will be an asterisk added by someone or other..

[Ryoshishokunin 301]

In commemoration of a large number of players, I have submitted my half-assed entry.  

[Ganzohnesushi 518]

On 29/01/2025 at 15:11, Ganzohnesushi said:

Seems I'm a privileged person. Until now I almost always received a confirmation e-mail. Also for Toto, ISP and many others. Now I hope I have not jinxed myself.

Ganzohnesushi 

Confirmation e-mail received as usual 

Ganzohnesushi 

[Bunbukuchagama 732]

12 minutes ago, Ganzohnesushi said:

Confirmation e-mail received as usual 

Ganzohnesushi 

Same here.

[Asashosakari 19,394]

An entry that feels like anything from yusho to makekoshi is possible...

Hoshoryu (O1w 12-3 Y)    Y    ---
Onosato (O2w 10-5)       O    Kotozakura (O1e 5-10)
Daieisho (Sw 11-4)       S    Oho (M3w 12-3 D)
Kirishima (M1w 11-4)     K    Abi (Ke 7-8)

Wakatakakage (Kw 7-8)    M1   Wakamotoharu (Se 6-9)
Gonoyama (M3e 8-7)       M2   Chiyoshoma (M5w 9-6)
Tobizaru (M2e 7-8)       M3   Takanosho (M1e 6-9)
Takayasu (M6e 8-7)       M4   Ichiyamamoto (M6w 8-7)
Ura (M4w 7-8)            M5   Hiradoumi (M5e 7-8)
Kinbozan (M14w 12-3 D)   M6   Shodai (M4e 6-9)
Atamifuji (M2w 5-10)     M7   Takerufuji (M11w 10-5)
Tamawashi (M10e 9-6)     M8   Oshoma (M9w 8-7)
Endo (M7e 6-9)           M9   Hakuoho (M15e 10-5)
Nishikigi (M12e 8-7)     M10  Shonannoumi (M13e 8-7)
Midorifuji (M11e 7-8)    M11  Meisei (M10w 6-9)
Takarafuji (M8w 5-10)    M12  Onokatsu (M12w 7-8)
Nishikifuji (M17e 9-6)   M13  Shishi (J4w 13-2 Y)
Churanoumi (M9e 4-11)    M14  Ryuden (J3e 11-4)
Aonishiki (J5w 12-3)     M15  Sadanoumi (J1e 9-6)
Asakoryu (J2w 9-6)       M16  Kotoshoho (M13w 5-10)
Mitakeumi (M7w 2-13)     M17  Tokihayate (M17w 7-8)
Shirokuma (J4e 9-6)      M18  ---

[Reonito 1,469]

6 minutes ago, Asashosakari said:

An entry that feels like anything from yusho to makekoshi is possible...

Differences from mine:

• Oho/Kirishima swapped (4 points) 
• I went Ura, Kinbozan, Ichiyamamoto, Hiradoumi (8 points!)
• Endo and Hakuoho swapped (2 points)
• Churanoumi behind Ryuden and Aonishiki (5 points)

 

[Gurowake 4,084]

4 hours ago, Asashosakari said:

An entry that feels like anything from yusho to makekoshi is possible...

I have similar confidence (or lack thereof) in my choices, as I went perhaps a bit too strongly by numbers alone.  I feel like if I make adjustments for joi rikishi in anything but breaking ties, it'll be the one time where they decide to ignore that factor, because I have a memory of that happening to me before.

Hoshoryu(12-3 Ow)	Y
Onosato(10-5 Ow)	O	Kotozakura(5-10 Oe)
Daieisho(11-4 Sw)	S	Kirishima(11-4 M1w)
Oho(12-3 M3w)	K	Abi(7-8 Ke)
Wakatakakage(7-8 Kw)	M1	Wakamotoharu(6-9 Se)
Gonoyama(8-7 M3e)	M2	Chiyoshoma(9-6 M5w)
Tobizaru(7-8 M2e)	M3	Takanosho(6-9 M1e)
Takayasu(8-7 M6e)	M4	Ura(7-8 M4w)
Kinbozan(12-3 M14w)	M5	Ichiyamamoto(8-7 M6w)
Hiradoumi(7-8 M5e)	M6	Takerufuji(10-5 M11w)
Shodai(6-9 M4e)	M7	Atamifuji(5-10 M2w)
Tamawashi(9-6 M10e)	M8	Oshoma(8-7 M9w)
Hakuoho(10-5 M15e)	M9	Endo(6-9 M7e)
Nishikigi(8-7 M12e)	M10	Shonannoumi(8-7 M13e)
Midorifuji(7-8 M11e)	M11	Takarafuji(5-10 M8w)
Meisei(6-9 M10w)	M12	Onokatsu(7-8 M12w)
Nishikifuji(9-6 M17e)	M13	Shishi(13-2 J4w)
Ryuden(11-4 J3e)	M14	Aonishiki(12-3 J5w)
Churanoumi(4-11 M9e)	M15	Sadanoumi(9-6 J1e)
Asakoryu(9-6 J2w)	M16	Mitakeumi(2-13 M7w)
Kotoshoho(5-10 M13w)	M17	Tokihayate(7-8 M17w)
Shirokuma(9-6 J4e)	M18

Edited February 18 by Gurowake

[Gurowake 4,084]

3 hours ago, Reonito said:

Differences from mine:

• Oho/Kirishima swapped (4 points) 
• I went Ura, Kinbozan, Ichiyamamoto, Hiradoumi (8 points!)
• Endo and Hakuoho swapped (2 points)
• Churanoumi behind Ryuden and Aonishiki (5 points)

 

I think I have those same differences, plus swapping Meisei and Takarafuji, Mitakeumi and Kotoshoho, and putting Takerufuji ahead of Shodai and Atamifuji, a choice that I feel is likely wrong but I'm too much of a slave to the numbers.  

Edited February 18 by Gurowake

[Tigerboy1966 1,449]

Hoshoryu(12-3 Ow)	Y
Onosato(10-5 Ow)	O	Kotozakura(5-10 Oe)
Daieisho(11-4 Sw)	S	Kirishima(11-4 M1w)
Oho(12-3 M3w)	K	Abi(7-8 Ke)
Wakatakakage(7-8 Kw)	M1	Wakamotoharu(6-9 Se)
Gonoyama(8-7 M3e)	M2	Chiyoshoma(9-6 M5w)
Tobizaru(7-8 M2e)	M3	Takanosho(6-9 M1e)
Takayasu(8-7 M6e)	M4	Ichiyamamoto(8-7 M6w)
Kinbozan(12-3 M14w)	M5	Ura(7-8 M4w)
Hiradoumi(7-8 M5e)	M6	Takerufuji(10-5 M11w)
Shodai(6-9 M4e)	M7	Tamawashi(9-6 M10e)
Atamifuji(5-10 M2w)	M8	Oshoma(8-7 M9w)
Endo(6-9 M7e)	M9	Hakuoho(10-5 M15e)
Nishikigi(8-7 M12e)	M10	Shonannoumi(8-7 M13e)
Midorifuji(7-8 M11e)	M11	Takarafuji(5-10 M8w)
Meisei(6-9 M10w)	M12	Onokatsu(7-8 M12w)
Nishikifuji(9-6 M17e)	M13	Shishi(13-2 J4w)
Ryuden(11-4 J3e)	M14	Aonishiki(12-3 J5w)
Churanoumi(4-11 M9e)	M15	Sadanoumi(9-6 J1e)
Asakoryu(9-6 J2w)	M16	Mitakeumi(2-13 M7w)
Kotoshoho(5-10 M13w)	M17	Tokihayate(7-8 M17w)
Shirokuma(9-6 J4e)	M18

[Andonishiki 170]

Didn’t play this time…

too many players >> chances to get kk are too slim now :)

just kiddin

the more players, the more super banzuke points, the top 42 will receive, right ?

as some of us really care A LOT about that, it somehow does matter

Edited Thursday at 13:42 by Andonishiki
But grandpa over in Kyushu DID play, yeah ! Forza Wakudoyama