XBoxSports.com  




View Poll Results: Should future tournaments use the "A" and "B" pool method?  
Yes  Distinguish an "A" and "B" pool  10  45.45%  
No  Keep the structure and matchups of the tournament random  12  54.55%  
Voters: 22. You may not vote on this poll 

Thread Tools  Display Modes 
Jun. 9th, '17, 2:56 pm  #11  
Frozen Water Sports

For this tournament, if we were to implement this, each player would choose their pool. Then, going forward, each player would eventually be designated as either an "A" or "B", based upon their performance, always with the opportunity to move up or down in future tourneys, if performance warrants.



Jun. 9th, '17, 5:13 pm  #12  
Somewhat of a Cone

Could I offer an alternative solution if we're considering pooling? You mentioned there's no "overall" rating for the players, but with a small field such as this, we can easily achieve that result with a functionality called an ELO Rating. Instead of a tournament at the start, you could set up a tablestyle setup and we can start calculating ELO Ratings that would maintain through the tournaments as we go. That way, we can see the true sense of talent and if you guys are sold on a bracket tournament option over a table format, we can start tabulating results onthefly, and leave it as one pool for now, and expand later (or keep it the same).
Also, I'd suggest a 3rd place game as well for those who made it to the final 4 but got crushed. No bias there. Last edited by F5 Penguin; Jun. 9th, '17 at 5:20 pm. 


Jun. 9th, '17, 5:54 pm  #13  
Baron Von Versus Hands

What is ELO and how is it calculated?




Jun. 9th, '17, 6:50 pm  #14  
Somewhat of a Cone

The longform answer can be best described here: https://en.wikipedia.org/wiki/Elo_rating_system
However, shortform I can assist with. Each player is assigned a rating upon entering the association based on a set number of games to decide his initial rating. For us, I'd probably just start calculating based on the random bracket draws. Each game will change the player's rating based on if he wins, loses, or draws. The ELO rating has been migrated a bunch of times to be used for other sports since then and is popular among ladder systems for online gaming. Essentially everyone would start at 1500 and I can do some calculations to set the base scores for those who have already played. Further, it should be noted that winning 50 would be weighted slightly more than a 21 game. It's something we could apply now and work it out as more games are played if this tournament becomes popular and more games are played. I could go into a lot more detail but that's the gist of it. Just another option I figured I'd throw out there. It'll give us a good sense of performance and power and those sandbagging intentionally to be placed in a lower bracket could be identified. I may whip up the current calculations if I have some time in a bit. Edit Add: The very quick formula would be: 1 / 10(R[pre]R[opp])/400+1 Where R[opp] is the opponent's rating. I can also add in coefficients for the finals and the score if we wanted, but that formula would just calculate a quick win/loss points bump/reduction for a clear, concise rating system. Additional edit: There's more to formulas and crap but I can work out whatever formula we want and it's easy I just don't feel like going into detail right now because it is currently a theory. Last edited by F5 Penguin; Jun. 9th, '17 at 8:45 pm. 


Jun. 9th, '17, 10:47 pm  #15  
Somewhat of a Cone

So... I did a quick calculation after my broadcast tonight. This is done with the following parameters:
1 All users start at 1500. 1500 represents an average XBS Versus player. 2 All series were counted as ONE win. Ex: if you win 2 games to 1, 2 games to 0 or 4 games to 0, it counts as 1 win. 3 No goal differentials were calculated. 4 Games were rated in sequence. Obviously it's the simplest formula and I can advance from there but I just wanted you to see what it would look like using the base set of data we currently have. 


Jun. 10th, '17, 5:52 am  #16  
Senior Member

Last place, baybay!



Jun. 10th, '17, 6:48 am  #17  
Senior Member

You know what they say Kabbot if your not 1st your last.



Jun. 10th, '17, 8:14 am  #18  
Senior Member

True. So I guess everyone is last.



Jun. 10th, '17, 8:16 am  #19  
Somewhat of a Cone

LOL. For this case, yes.
I know you're joking but I feel like explaining stuff below Unfortunately I didn't calculate total games played and weight it with goals scored. Also, through some research a year ago, home ice has a .006% higher chance of winning versus games or something close to that (I need to pull those numbers again), and that's not calculated either. So if I did that and bring in all the other factors, you might not be last place. Currently, everyone who lost in the 1st round are technically tied for last. Edit: And then there's the whole "expected win" probability that needs to be factored in as well. That says if Gretz plays Kabbott, based on current ratings, Kabbott should win by x percentage. Last edited by F5 Penguin; Jun. 10th, '17 at 8:54 am. 


Jun. 10th, '17, 9:07 am  #20  
Member

I'll participate regardless....


