Board Game Arena

FrankJones wrote: ↑30 September 2026, 20:03 But his accomplishments disprove the assertion that even the worst player will beat the best players 35-40% of the time.

People think the same about backgammon. Incorrectly.

yodamus wrote: ↑30 September 2026, 20:23 Is this the same Stu Unger that was top pro poker player ? Memory goes a long ways into being better at games than others.

Yes. The greatest Gin Rummy player ever, without question, and arguably the best poker player ever.

Fascinating person and fascinating (though tragic) life.

Meeplelowda wrote: ↑30 September 2026, 20:28

yodamus wrote: ↑30 September 2026, 20:23 Is this the same Stu Unger that was top pro poker player ? Memory goes a long ways into being better at games than others.

Yes, but he was playing gin rummy, at a pretty young age if memory serves, before he played poker. And gin rummy is absolutely a game in which memory is an asset.

It's way more than just that though. Give a typical gin rummy player a game log and that player till wouldn't be able to do whatever god-like algorithm Stu Ungar's brain performed to know his opponent's entire hand.

I doubt even the best gin rummy players here could identify every card in the opponent's hand after 3 discards.

Meeplelowda wrote: ↑30 September 2026, 19:59

FrankJones wrote: ↑30 September 2026, 18:26 How did Stu Ungar manage to win almost 100% of the time? (At gin Rummy.)

A combination of a freakishly good memory and being really good at the game? The latter may seem to be a byproduct of the former to a degree, but you need to know how to use what you remember correctly, and that's what I mean by being really good at the game. I estimate a one in a billion chance you haven't already seen this, but just in case I recommend "High Roller: The Stu Ungar Story."

It's like those people who can look at a city landscape and reproduce it in every detail. Those people are not playing in the same category.

So every other top player may be able to achieve a 70% winrate at best, but there's one person out there who has this special ability that allows them to win almost 100% of the time.

tbhp wrote: ↑30 September 2026, 18:08

hibachi wrote: ↑30 September 2026, 17:41 In a game where top winners are winning at 60%, you shouldn't lose 3x the Arena points that you gain for a win...even against someone you are significantly higher ranked...because thee worst player is still 35-40% to win against the VERY best player.

The problem isn't the number of points you lose (though in your example it would be 2x, not 3x), it's that at such a slim difference between the best and the worst player, your Elo is always going to be hugely impacted by the luck of your recent games.

The only solution would be to create a non-Elo system, but seeing as Arena makes everyone play against everyone, what would that system be?

You could simply list the win percentages of each player (with maybe a requisite that you have to play like 100 games to appear in the ranking). But given that players are randomly assigned opponents, there would be zero guarantee that the games two players get are fair to compare.

No it's actually very commonly 3x in Arena Points. I played the same player back to back last night...lost one (-32 points), and won one (+11 points). For what it's worth I'm not saying he's a bad player...but his ELO was low 100's when we played.

I don't know the best solution - I'm pointing out the current system is less than ideal for games like this. There has to be a way to lessen the swings for games that are more luck based.

FrankJones wrote: ↑30 September 2026, 18:26

hibachi wrote: ↑30 September 2026, 17:41 ...because thee worst player is still 35-40% to win against the VERY best player.

If the worst player really can win 40% of the time against the best player, then no system is likely to ensure that the best players win an arena season.

How did Stu Ungar manage to win almost 100% of the time? (At gin Rummy.)

He didn't win anywhere close to 100% lol, unless he was playing total fish every time that were clueless.

FrankJones wrote: ↑30 September 2026, 21:25

yodamus wrote: ↑30 September 2026, 20:23 Is this the same Stu Unger that was top pro poker player ? Memory goes a long ways into being better at games than others.

Yes. The greatest Gin Rummy player ever, without question, and arguably the best poker player ever.

Fascinating person and fascinating (though tragic) life.

I know nothing about gin rummy, but the poker stuff is only based off context of the time. Obviously, Ungar was a great poker player for his time, but If you measured poker by ELO, the best poker players of Stu Ungar's days wouldn't even beat .5/1 online today. Poker basically completely changed/got revolutionised by online poker.

As for gin rummy, because I don't even know how it's played it's hard to comment directly, but if we're saying someone could win near 100% of games in the 80s but today a top player can only win 60% it seems suspicious to me that something else is going on. Like maybe nobody is playing gin rummy, so it's not really a game becoming solved over time, but I would expect pretty much any game the standard to have improved in the last 40 years, especially with the advent of computers being such a useful tool at helping people improve. The digital world basically makes it impossible for you to be that much better than everyone else at the game, because people can easily learn from your play/techniques in a way that simply wasn't possible in the 80s.

hibachi wrote: ↑30 September 2026, 21:39

FrankJones wrote: ↑30 September 2026, 18:26

hibachi wrote: ↑30 September 2026, 17:41 ...because thee worst player is still 35-40% to win against the VERY best player.

If the worst player really can win 40% of the time against the best player, then no system is likely to ensure that the best players win an arena season.

How did Stu Ungar manage to win almost 100% of the time? (At gin Rummy.)

He didn't win anywhere close to 100% lol, unless he was playing total fish every time that were clueless.

He famously beat the top known player 86 games to 0.

RazorOz wrote: ↑30 September 2026, 22:25

As for gin rummy, because I don't even know how it's played it's hard to comment directly, but if we're saying someone could win near 100% of games in the 80s but today a top player can only win 60% it seems suspicious to me that something else is going on. Like maybe nobody is playing gin rummy, so it's not really a game becoming solved over time, but I would expect pretty much any game the standard to have improved in the last 40 years, especially with the advent of computers being such a useful tool at helping people improve. The digital world basically makes it impossible for you to be that much better than everyone else at the game, because people can easily learn from your play/techniques in a way that simply wasn't possible in the 80s.

Stu Ungar was willing to make an even-money bet that he could identify his opponent's entire hand after 3 discards. That's a skill nobody else possessed, then or now. The era has nothing to do with it. Unlike poker, with betting and ICM strategies that have completely revolution the game, that's not true of Gin Rummy. Gin Rummy is still just Gin Rummy.

The top players of today aren't Stu Ungar, that's perhaps why they only win 60% of the time.

I also call into question to accuracy of the statement that the "top" players today can only win 60% of the time.

hibachi wrote: ↑30 September 2026, 21:36

No it's actually very commonly 3x in Arena Points. I played the same player back to back last night...lost one (-32 points), and won one (+11 points). For what it's worth I'm not saying he's a bad player...but his ELO was low 100's when we played.

I agree about the 3x. I play an even higher variance card game, and if I were to randomly play arena, I would struggle to win often enough to offset the losses. For this game, the arena winner is determined by who gets lucky at the end of the season.