BG, no one is overlooking great performance. However, our league rules are that to win an annual award, you have to attend more than half the games. The only "penalty" you would suffer is not winning a paper award. You wouldn't be excluded from anything except those awards, which are mostly about bragging rights.
If it doesn't mean anything, why have a league in the first place?
Bloody, I don't have an explanation for the different scores. I sent you two sheets, and the second one did have a correction in it. I asked you to send me the input data so I could copy it exactly. If I didn't input it exactly, I'd get different results. Neither of us would know that without looking at the input data.
You misunderstood my comment about what we measure. The 10 events you used as an example include factors our system doesn't measure. I'd say up front that our system is not designed to measure what we are trying to measure in that format. We'd see those as differences that should be measured. To me, those are actually games being played by different rules, so I'd be reluctant to try to measure them together using our system because they are so different. Our system is designed to measure games played with the same structure and rules.
You clearly don't understand points systems, else you'd understand that when comparing points system A, which incorporates buy-ins and rebuys, to points system B, which doesn't incorporates buy-ins and rebuys, you are making a fair comparison when you take out the buy-ins and rebuys as variables from system A. You can't then hide behind the fact that system B doesn't incorporates buy-ins and rebuys, because now neither does system A.
We're comparing the same rankings and the same players. How they got those rankings is beside the point for the purpose of comparisons.
If we knew there would be different entry amounts and a mix of freeze-outs and re-buys, we'd build our system to measure those things, and I'm pretty sure it would make a difference. While you did state you were measuring systems based on what they did, the example you provided is not one I'd have our current system measure. Our system was specifically designed for a "league" where every game has the identical structure and rules. If we were measuring something like yours, we'd try to adjust our system to measure what we consider relevant factors. There are reasons for that.
- Some players will not play in re-buy events. Our system does reward greater attendance, a point you dispute but I'll address. Players who refused to play in certain events would mean they really aren't completely part of the league, though they might play in some games and their stats would be counted. However, they aren't committed to the entire season, only those events that meet their personal criteria.
Besides the point. Rebuys and buy-ins were removed from the comparisons.
- Some players will not play if the entry fee is above their budget. They might play in $40 but not $60.
Besides the point. Rebuys and buy-ins were removed from the comparisons.
- Some of those might play in re-buys but would count one re-buy as part of the entry fee would be priced out. They might play in $100, but not a $60 with a re-buy since they'd see that as a $120 entry fee.
Besides the point. Rebuys and buy-ins were removed from the comparisons.
Collectively, these factors mean there are possibly 3 groups of players who wouldn't play in all events. We'd know some players are not really going to compete for anything because they wouldn't come to enough games. One could be a great freeze-out player but wouldn't qualify for an award for lack of attendance. We might recognize though that when that player comes, he's among the best players though.
What does this have to do with points scored? "We think you're one of the best players, we're just not allowing your excellent performance to be a part of our league".
I think we, like most leagues, have players who come who won't compete for annual awards because their attendance doesn't allow them to compete, but what makes our format attractive to them is that each game itself stands on its own. There is no payment that goes toward another game -- 100% of the entry fees are prize money in that game. Thus, if BG came and won 5 games, he cleans up when it comes to money, and certainly would score well in our system. But at the end of the year, he wouldn't be ranked amount our top 10 or awards because of low attendance.
Where does this come from? Nowhere is there any mention in any of the posts that winnings aren't paid out to the tournament winners, or of any money going towards another game. Not to mention it has absolutely zero bearing on points calculation.
Poker isn't like team sports where every game is played under identical rules and every team plays the same number of games to make it to the playoffs. So I think it's hard to compare players over 3 types of games, especially in a short season. But if you were going to do that, a system that accounts for those differences to me would be essential to the comparison. I think I mentioned that I looked at Dr. Neau's system. Two reasons I rejected it was that we had a consistent format and it seemed to me his was designed to measure smaller games than we had.
Absolute horseradish. Dr. Neau's points system can be applied to any number of players, unlike your system which assigns points to only the top 7 finishes regardless if you started with 7 players (top 100%) or 700 (top 1%) which limits your system to fieldsizes where "7" would make sense, and if you enter the same buy-in for every game, and "0" for rebuys and add-ons Dr. Neau's formula works perfectly (or you can use the simplified version listed in the OP which leaves those values out entirely).
Our #2 player last year was a guy I hadn't met and he didn't know about our group until the 4th game of the year. He had no chance of playing in the first 3 events. He made 8 of the 9 he could have attended. By points he would have finished 6th. I doubt another system would have rated him much higher than that. But his performance in those 8 games was second only to the top guy. He was the Bertrand of your example in a sense.
Did you test this on any other points system, or is that just your assumption? Did you take into consideration that you can count only the top X results?
Bertrand missed 3 of the first 6 games, then attended the rest. His performance in those 7 games was better overall than Sammy and Phil, but since they attended more games, they did accumulate more points. No one could intelligently argue that any of the three are clearly better players. Anyone could intelligently argue that any of them could have finished first under the right system. Where they finish depends on what that system rewards.
Discussed and debunked before. You can count the top X results only. Problem solved.
Here is the risk in our system of a player (A) not coming because he thinks he is too far ahead. A could sit out, and one of the players behind him (B) could pass him with a good performance. By sitting out, A doesn't get the participation points. Even if A got only participation points, that does help A's score, not hurt. How, A's per game performance might go down, but B coming and doing well would improve B's per game performance and it might be enough to pass A.
Categorically untrue, as demonstrated by Sammy sitting out after winning the first game. (And don't start with the 7 game minimum, the same logic applies there as well if you've built up a lead over the first 7 games).
This specific issue is one we were concerned about. In 2014, the overall leader had a commanding, but not insurmountable lead, going into the last tournament. It turned out that both people behind him did perform well. Had he not showed up, he could have fallen in the standings precisely because every player's overall performance is compared to overall performance of every other player. A good comparison would be two baseball players chasing the league batting title, which is based on batting average. The one who is slightly ahead might decide to sit the last day of the season so his average doesn't go down, especially if he's facing a good pitcher who he hasn't done well against. But the risk is the player right behind him has a great day and passes him. I've played with our system enough to know sitting out doesn't really help a player. But I also know that attending 10 games instead of 7 doesn't make one a better player.
Debunked before.
If BG came to only 5 of our 12 games and won them all, I think it's highly likely he was the best player in the group! But league rules might not allow him the same benefits as others who played more, and it's possible that in a points only system, he wouldn't finish first, or maybe not even anywhere near it. So this gets back to what exactly you are trying to measure.
At least BG would be a contender for the top prize in the other points systems.
Here is a simple concept. Systems that reward most those who come the most, regardless of their skill, ignore a higher level of performance for those who attend less unless their performance was enough higher on its own to score higher. The top player in that system is not necessarily the best player. The top player might have won due to circumstances of another player -- the guy who had to sit out a game due to a death in the family, illness, etc. It puts a premium on attendance. That's not wrong, but that is not what we wanted to measure.
Described before, and options offered before.
When I was looking for a system, I couldn't find one that didn't base things primarily on attendance. In every system I looked at, players who attended more games accumulated more points. That's not wrong, but I felt other things should count as well.
Then you didn't look hard enough, or didn't know of or thought about the possibility of counting only the top X scores.
Now if I'm hosting, an attendance-based scoring system always helps me because I'm the one guy they can't play without. While I'm highly competitive and would never do this, suppose after 11 tournaments, I was ahead and wanted to win. I could dream up a reason to cancel the last game of the year. Should that decide who wins awards? What if those awards are worth real money? Should that decide who gets to the final game, especially if players who don't qualify are helping fund that game?
So you're saying that the other points systems don't work because you don't trust the persons hosting them? And again, where is this real money award coming from? And how does that have anything to do with how the points are calculated? You think a league has any viability for a next season if the host pulled a stunt like that? And how is that not true for your system as well if you're leading in the points after game 11?
I don't think I need to mention again that you can count only the top X results...
A league that rotates locations is less dependent on a host and thus would be more difficult to manipulate.
The problem I attempted to solve is how do you determine what factors measure skill and how you weigh those factors. I'm very interested in improving our system. It does measure more things than any other system I've seen, and if it doesn't generate different results, it's pointless to measure those things. We could add things to it if we could identify them. Different entry fees and formats could be added, but both are irrelevant to our format. That's why we don't measure them.
And neither do any of the other systems if there are no rebuys and if the buy-in is the same for all games.
Instead you award escalating points for the top 7 only (an arbitrary number), and you award points for making the final table (even if you're only playing at one table to begin with), and points for winning money (which only the players in the top 7 can win), and for finishing first (which already gets you the most points), and you award points for receiving points.
Most (or all?) of these you're calculating twice, like having a box of 20 apples of which 12 are red, and a box of 10 apples of which 2 are red, and calculating the percentage of red apples as (60% + 20%) / 2 = 40%, and 14/30 = 46.7%, and then averaging them out to get to 43.3% red apples.