Cactus Game Design Message Boards

Haha, I knew I'd be high on that list - Despite being 2-2 against the 'Top 20' I'm also 2-2 against the 'Top 6'

I beat 3rd and 5th, and lost to 1st and 6th, I didn't play 4th, and I'm player 2.

What does that list look like if you factor in the past two nationals? Am I top? or does somebody still pass me?

Chris played 9 of the top 20 people. If that's not a strong schedule, I don't know what is.

here is the 2012 Only Top 25.

1. Martin Miller                   2010
6. Chris Ericson                 1824
11. Connor Magras           1747
12. James Roepke             1736
14. James Courtney            1685
15. Andy Stanley              1681
18. Mark Underwood   1654
22. Ben Michaliszyn         1623

This is another interesting way of ranking things.  I am curious though why I am ranked down at 18 in this system compared to finishing 12th at the actual tournament.  My strength of schedule should have been pretty high too, considering that I played all of these people from the list, and considering that I won against Magras, Roepke, Stanley, and Michaliszyn, yet am ranked lower than 3 of those guys.

It also matters who you lost to, and how they fared against the others.

If you can relate to college football: If Michigan State beats Michigan, it does not mean that they end up higher than Michigan in the National Rankings. All of the other games matter too. But timing is important, too, as Alex mentioned. Losing the first game of the year, even to a lowly opponent, is better than losing just before the bowl games. Virginia Tech proved this even after losing their opener to Richmond.

Glicko awards points based on the comparison at the time of the game, rather than based in the final finish of the player you beat or lost to.

That would be a problem with this method then, because a first round game against Gabe or Martin is NOT the same as a first round game against a RLK.  For this method to be helpful, people would have to have their Glicko points carry over from year to year.

Personally, I think it would be fun to use the method employed by professional soccer: pool play at the beginning followed by knockout stages.  The ranking of players could be used to spread out all the top players into the different pools/groups.

I'm not sure if you're being serious or sarcastic, but this is the basic idea that I'm also throwing out as being a good way of ensuring final round match up top players, while at the same time giving everyone a chance to keep moving up all the way to the last match if they play well enough.

Quote from: Alex_Olijar on September 04, 2012, 08:59:09 AM

Glicko awards points based on the comparison at the time of the game, rather than based in the final finish of the player you beat or lost to.

That would be a problem with this method then, because a first round game against Gabe or Martin is NOT the same as a first round game against a RLK.  For this method to be helpful, people would have to have their Glicko points carry over from year to year.

Quote from: STAMP on September 04, 2012, 12:44:00 PM

Personally, I think it would be fun to use the method employed by professional soccer: pool play at the beginning followed by knockout stages.  The ranking of players could be used to spread out all the top players into the different pools/groups.

I'm not sure if you're being serious or sarcastic, but this is the basic idea that I'm also throwing out as being a good way of ensuring final round match up top players, while at the same time giving everyone a chance to keep moving up all the way to the last match if they play well enough.

I'm serious about the idea.  My sarcasm comes in the last statement which is to poke jabs at those that like to argue about specific cards. 

Quote from: Alex_Olijar on September 04, 2012, 03:42:13 PM

If you pool players, it makes it significantly harder to average to bad players who have a good tournament to get out of their pool, whereas in top cut they would be able to qualify for the top cut much easier than qualifying for a knockout stage.

I thought the whole idea of top-cut was to do a better job of ensuring that the best players end up having to play each other to win the tournament.  If so, then this comment doesn't make sense, because you seem to be agreeing with me that a top-cut system REALLY doesn't do that, but in fact does the opposite by making it EASIER for players that are NOT as good to make it into the top places of a tournament.  As long as they can be in the top 16 after a smaller number of rounds, then they ensure that they stay up there, and keep other better players below them from moving up in the rankings.

As mentioned in the other thread, the goal to is to isolate the best players in that tournament, not the best players at large. Sometimes Gabe isn't in the top cut because his performance in that tournament doesn't justify it, and that's ok.

Meanwhile, the proposed pool system would create even more inequality. There is no way you could ever create even pools. So what happens when Gabe gets an easy pool while RDT, yourself, and I have to gut it out for two spots in a separate pool? Is that fair?

There is no perfect fairness solution. However, top cut ensures that the best players on a given day as proven by results that day are forced to play each other, and avoids the issue we currently have with complex tie breakers.

You managed the skip the part of that sentence which explains exactly what top cut is good for.

Quote from: Prof Underwood on September 04, 2012, 03:37:22 PM

That would be a problem with this method then, because a first round game against Gabe or Martin is NOT the same as a first round game against a RLK.  For this method to be helpful, people would have to have their Glicko points carry over from year to year.

In fairness I only published the 2012 Only Top-25 to help make RDT feel better about his performance. The Current and Highest lists given before did roll over ratings from previous years. On those you are #81 on the Current list and your 2007 performance earned you #71 on the Highest list.

I wasn't worried about feeling better I know how good/not good of a player I am. I don't need a list to tell me differently

You managed the skip the part of that sentence which explains exactly what top cut is good for.

I didn't skip that by accident.  I left it out because I'm not convinced that it is true that top cut matches up the best players on a particular day.  It really just matches up the players who win the most in a few early rounds against unequal competition.  That's why I'm not seeing it as improving the overall fairness of a tournament compared to swiss.  However, I do concede that it would be helpful for eliminating confusing ties.

MJB, where I am sitting in all time glicko? I'd imagine low, but I'm curious.

30 points or so above RDT. 

Quote from: Red Dragon Thorn on September 03, 2012, 09:47:23 PM

Clearly I'm not that good of a player I'm not on any of these lists - This takes so much pressure off

You are in 91st place currently and 100th on the highest ever list.  (Unfortunately, the way I did the Glicko--complete carry-over of all results with no time-decay--will really punish someone who wasn't a phenom from the start.)

BTW, you should listen to Olijar he is 30 spots or so above you on both lists.

Your current and highest rating is 1760 which places you 60 on the Current/63 on the Highest.

Actually you are 1760 +- 79 and RDT is 1707 +- 71 (current).  You are tied within the deviation.

Hey guys, let's keep this topic strictly about looking at the rankings, not their implications for/against top cut. We already have a topic for that (where the same basic things are being rehashed), so if you want to discuss those implications, move it over there please. Kirk made this topic separate for a reason, I'll ask that we respect that.

Thanks Chris. I keep forgetting which thread I am reading...

Kirk