Ideas to improve the Guru system

[tbhp]

I can pretty much tell you that is how Arena works. I have won Heat Arena 11 seasons in a row, it's not because I am going on a lucky streak every season. Heat is a game with some luck, but still a lot more skill-based than high luck games. If you have the highest win rate at a game like that, you will win in Arena pretty much as long as you play enough games to reach your level, that's why a lot of the same players repeatedly win in Arena. I know from experience when I have tried playing significantly more games to try and increase my Arena ranking it doesn't work, I just end up going round in circles.

See the part I added to my last post. Your case is very specific, because if you won Arena so many times it clearly suggests you are in a different category from everyone else. So in your personal case, you don't need to extend your luck, you just need to reach your expected score. Then it's others who have to extend their luck to catch up to you. But given the gap may be quite large, the odds of getting enough consecutive wins for them to do so may be very low, which is why it has never happened.

In most games though, it's going to be unlikely that someone is clearly better than everyone else. Usually it's more like there's a group of top players who are roughly the same level, so the number of games played and the streaks achieved will have a large influence on which one of those players will win.

[tbhp]

I know from experience when I have tried playing significantly more games to try and increase my Arena ranking it doesn't work, I just end up going round in circles.

Regarding that part, what I mean is that if your true level is say at 2000 points, any game you win starting from that number is going to add more points. If you win one more game, maybe you'll be at 2010. If you win 20 more games in a row, maybe you'll be at 2100.

To you it probably doesn't matter, because other players have always been behind you, so you've never had to extend your luck to try to get much higher than your normal number.

But now imagine there is another player who is just as good as you. You'd both be around 2000, so you'd both be trying to play a ton more games to try to increase this number as high as you can. At some point one of you may manage to get to a number that seems unrealistic for the other to ever reach, and so they decide to stop playing.

That's what Arena is like for most people. For everyone who is not so far above everyone that they just need to reach their expected score, the final race is all about putting out games and hoping for win streaks.

Your perspective is as if Magnus Carlsen said "being the world Elo leader is not about luck streak, you just need to be at your normal level". That statement only holds true for him, since he is notably better than every other player. For everyone else, reaching the world top Elo would require a major win streak to surpass Carlsen's Elo.

[ChiefPointThief]

I used to be wary about playing real time, because I thought it wouldn't give me the necessary time to really be sure about my moves. I basically went for the quality over quantity approach. But as it turns out, players in general may not be as strong in real time, so there is actually no guarantee I would get a better winrate playing turn based.

Are you saying players in a specific game who play turn based are better or overall? & this is just one reason why the idea of arena competitively is flawed. There are actually 3 different player pools within each game (real time, slow turn based, fast turn based). Then people’s ratings for each season are being compared to entirely different player bases.

As you mentioned in a specific game 70% of the top players can be turn based players and that can be flipped for another game. So depending on the mode you play the pool can be less challenging. This makes me want to go back and post in that thread about how tournaments should have more importance

[tbhp]

I used to be wary about playing real time, because I thought it wouldn't give me the necessary time to really be sure about my moves. I basically went for the quality over quantity approach. But as it turns out, players in general may not be as strong in real time, so there is actually no guarantee I would get a better winrate playing turn based.

Are you saying players in a specific game who play turn based are better or overall? & this is just one reason why the idea of arena competitively is flawed. There are actually 3 different player pools within each game (real time, slow turn based, fast turn based). Then people’s ratings for each season are being compared to entirely different player bases.

As you mentioned in a specific game 70% of the top players can be turn based players and that can be flipped for another game. So depending on the mode you play the pool can be less challenging. This makes me want to go back and post in that thread about how tournaments should have more importance

It is flawed, it doesn't make much sense to compare players who play different speeds within the same competition. That's why I think it might be good if the time setting was decided by the gurus too. But then some people might complain that it's not inclusive of people who only play real time, or only play turn based. But picking an Arena setting is not inclusive in the first place, as you are alienating all the people who don't like the picked format.

I wonder what should be the takeaway from it though. I mean, would the people who do best in real time do equally as well in turn based? Or are those just totally incomparable things? I know players who do great in real time Arena but don't have the same success rate in tournaments. It seems confusing to me because if they can do so well with little time to think, why aren't they able to do better in tournaments where they have days to think?

Then again, there's also people who are able to get to the top playing Arena in turn based, but don't do that well in tournaments despite them being turn based too. So, I guess there is no consistent logic to those things, and some settings may be better adapted to certain players for reasons that are hard to pinpoint.

[ChiefPointThief]

You can finesse a higher score in arena and some of the ways to do that don’t exist in tournaments so I can see a player not having the same results in tourneys as arena. (Some of these topics are also game dependent).

About timed games being different (also game dependent) I agree. Example Barenpark (a game fitting puzzles into a grid) in tb I can map out way further ahead compared to real time. I have all the time in the world to do so. In real time I play slow or normal never fast because I would probably time out. Someone who can play w/ a fast clock and achieve the same results I would say is a better overall player.

[tbhp]

About timed games being different (also game dependent) I agree. Example Barenpark (a game fitting puzzles into a grid) in tb I can map out way further ahead compared to real time. I have all the time in the world to do so. In real time I play slow or normal never fast because I would probably time out. Someone who can play w/ a fast clock and achieve the same results I would say is a better overall player.

Yeah, in turn based Harmonies I'll usually think of different setups to accommodate for the different cards that can come out (especially if it's a high stakes game). In real time there's less time to do that, I go a bit more with intuition.

Still, high scores happen all the time in real time, but maybe not with the same reliability that I can in turn based. But it doesn't seem to matter, because the opposition tends to be inconsistent too, even top rated players.

[Ceaseless]

I used to be wary about playing real time, because I thought it wouldn't give me the necessary time to really be sure about my moves. I basically went for the quality over quantity approach. But as it turns out, players in general may not be as strong in real time, so there is actually no guarantee I would get a better winrate playing turn based.

Are you saying players in a specific game who play turn based are better or overall? & this is just one reason why the idea of arena competitively is flawed. There are actually 3 different player pools within each game (real time, slow turn based, fast turn based). Then people’s ratings for each season are being compared to entirely different player bases.

As you mentioned in a specific game 70% of the top players can be turn based players and that can be flipped for another game. So depending on the mode you play the pool can be less challenging. This makes me want to go back and post in that thread about how tournaments should have more importance

It is flawed, it doesn't make much sense to compare players who play different speeds within the same competition. That's why I think it might be good if the time setting was decided by the gurus too. But then some people might complain that it's not inclusive of people who only play real time, or only play turn based. But picking an Arena setting is not inclusive in the first place, as you are alienating all the people who don't like the picked format.

I wonder what should be the takeaway from it though. I mean, would the people who do best in real time do equally as well in turn based? Or are those just totally incomparable things? I know players who do great in real time Arena but don't have the same success rate in tournaments. It seems confusing to me because if they can do so well with little time to think, why aren't they able to do better in tournaments where they have days to think?

Then again, there's also people who are able to get to the top playing Arena in turn based, but don't do that well in tournaments despite them being turn based too. So, I guess there is no consistent logic to those things, and some settings may be better adapted to certain players for reasons that are hard to pinpoint.

There's definitely a divide issue when it comes to games and variants. On Chess sites there are different ratings for different groups of time controls. Real time and turn-based play are different beasts that could justify different formats, though putting them to an either/or vote for Arena is unnecessarily alienating given a lot of players find one time control group or the other untenable for various reasons I don't see it being viable for BGA to get involved with. It's possible that having an Arena for each could work, though that might also be something that would be best to check game by game, I'm not sure how much BGA would want to deal with something of that nature. It's a different beast from game variants, where oftentimes there really is only one real leader for competitive play, and even among the times there are multiple contenders they hold enough weight that Arena votes may very well alternate as a result.

I mean, would the people who do best in real time do equally as well in turn based? Or are those just totally incomparable things? I know players who do great in real time Arena but don't have the same success rate in tournaments. It seems confusing to me because if they can do so well with little time to think, why aren't they able to do better in tournaments where they have days to think?

The skills that let you win in one time control give you some advantages in others, but not all skills have the same impact in each time control, so a player doing equally well is no guarantee. Given the right skill discrepancies, the advantage may even go to the other player. If I can beat someone in a game of Chess in 15-10 time control, that will probably correlate with my ability to beat them in 5-3, but some players will have speed skills that make it far from identical groups. Some critical moves may take a lot of time to sort out, which also means certain moves that can be matched but take up a lot of time to find can become more effective weapons in shorter time controls depending how players convert the difference.

Remember, both player get more time to think, not just one. So you can't say "they played so well with less time, why didn't they do better with more time?" Their opponent got more time too. So if Player A found the second-best move in 10 seconds, and could find the best move in 2 minutes, while player B has no good moves in 10 seconds, takes 10 minutes to find the second best move, and 1 hour for the best move, then if both players have 30 seconds for their move, player A is ahead, but if both players have 1 day, they are equal. What if it takes 1 month for someone to find the best move? Or 10 years? And we haven't gotten into the gap between the best and second-best moves. The clock is a weapon too.

Then again, there's also people who are able to get to the top playing Arena in turn based, but don't do that well in tournaments despite them being turn based too. So, I guess there is no consistent logic to those things, and some settings may be better adapted to certain players for reasons that are hard to pinpoint.

Tournaments are relatively tiny events, a better player still only wins X percentage of the time, if the dice don't fall in their favor they won't make it, while Arena offers far more games to help avoid a bad outlier messing you up. On the other hand, tournaments let you curate the player pool more, for better or worse. It can let matchups favor a player, or it can add a quality to the event that gives a higher-level representation than the low level grinding a blind pairing system throws at you. In the end the devil is in the details, and you have to look at the specifics in the hopes of getting a better idea of what went down.

[tbhp]

It's possible that having an Arena for each could work, though that might also be something that would be best to check game by game, I'm not sure how much BGA would want to deal with something of that nature.

There could be, but I'm not a fan of dividing things like that and ending up with two winners per season. I mean, who is the "real" winner then? It seems wonky and even though in sports such divisions exist (there's men/women, and multiple categories for the same sport), I think for Arena having a single winner is what makes most sense.

Tournaments are relatively tiny events, a better player still only wins X percentage of the time, if the dice don't fall in their favor they won't make it, while Arena offers far more games to help avoid a bad outlier messing you up.

It's true that tournaments open up the list of potential winners, and the best players can fail because things just didn't go in their favor this time. But that's looking at just one event. I think if you look at a player's general record, how often they won, reached finals, semi finals, also accounting for things like the number of players and their strength, this gives you a meaningful insight into how well a player does in tournaments. And some players just don't have the kind of record I would expect someone who does so well in Arena would have.

I think maybe the explanation is less about luck and is more of a psychological thing. A tournament setting is a different type of setting than going through games in Arena. There is a certain pressure associated with it, especially the further you advance in the tournament, where you whole success depends on not losing this one game.

It's kind of like in sports, once you've won the world cup once, it becomes easier to win it again. There can be a psychological barrier for the players or the teams who haven't known success yet, so no matter how good they are, they tend to sabotage themselves in the final stages.

[Respublik]

Maybe in practice a simple enough solution would be to ask people in good faith if they want to be a guru. Sure, some people might lie, but I think most of the people who play a lot of games but aren't interested in the Arena format would just say "yeah, I'm not interested, it's better for someone else to take my spot".

When you think about it, it's a bit weird that becoming a guru is something that is just forced on people. It should be possible to say no, if you feel it's not your role or like you aren't the target audience.

This has been in the suggestions basically ever since the system was introduced. Here: https://boardgamearena.com/bug?id=46208

[tbhp]

This has been in the suggestions basically ever since the system was introduced. Here: https://boardgamearena.com/bug?id=46208

The description of this report resonates with me. Of the 3 games I play competitively, they made me a guru for the 2 games where not only I am the least interested in being a guru, but also where I'm the less good at.

This goes to show that the idea to go by number of games played has limitations for several reasons. The first one has already been discussed, being that people are more likely to play a game a lot of if the current Arena format is one they enjoy, thus creating a self reinforcing circle. The second reason is that the number of games played is often a poor indicator of someone's skill at the game. It can even be inversely proportional to it, because if someone keeps failing in Arena, they are more likely to keep trying again and again.

They keep making me a guru in Castle Combo because I play this game the most, but the reason for that is because I'm the worst at it so I need to put out more games to get close to my objectives. So this is actually contradictory with the idea of being a guru.

In contrast, in Harmonies I've never played more than 100 games in a season, so I stand no chance of being selected when there's people who play hundreds of games, even though that's the game where I'd be most fit to be a guru.