Talk Fantasy Football

What about SoS based on sum of opponents' NAF win ratios?

@mubo:
Thanks for explaining it again. Now I got it
Yeah would agree to not follow in that direction.

I believe that the NFL Strength of Schedule is the sum of your opponent's records against other opponents. That is, your own games don't count toward your Strength of Schedule.

My objection to using Strength of Schedule as a tiebreaker is that it's not especially easy to calculate, at least not compared to common tiebreakers such as TD Differential. I like to look at a table and know where I stand, and know what I need to do. With a TD Diff tiebreaker, I can look at a table and know, "I need team A to lose, teams B and C to tie each other, and I need to win by at least three". Not all of that is under my control, but at least I can play my game (incredibly aggressively) and keep an eye on some other tables and know where I stand.

Although I can imagine a hilarious tournament based on Strength of Schedule, where table 1 ends their game with the same record and the entire tournament is decided by the last table's SoS contribution

Nippy Longskar wrote:What about SoS based on sum of opponents' NAF win ratios?

I wondered about something like this. Strength of schedule based on a single tournament's results is really saying 'strength based on how people performed this weekend'. Using NAF rankings or win percentages could provide a more 'absolute' base. I mean if those rankings can't on average tell us who is better then why do we maintain them?

As an add on question would you use a coaches overall ranking/win percent or the numbers for the race they were using that weekend?

@Rolo that's a really interesting point about SoS that I hadn't considered- that's it good if coaches know the tiebreaker situation going into their final game. Coaches would surely feel better about any tiebreaker if going into game 6 they knew the implications. I mean if you try for that 4-0 win but Draw 3-3 you know that those extra risks didn't pay off but you feel ok about taking them.

Greshvakk wrote:

Nippy Longskar wrote:What about SoS based on sum of opponents' NAF win ratios?

I wondered about something like this. Strength of schedule based on a single tournament's results is really saying 'strength based on how people performed this weekend'. Using NAF rankings or win percentages could provide a more 'absolute' base. I mean if those rankings can't on average tell us who is better then why do we maintain them?

As an add on question would you use a coaches overall ranking/win percent or the numbers for the race they were using that weekend?

I massively improve my game and pull it out of the bag and get 510 at a weekend. Only one other coach gets the same result. But because he's played better opponents in the past he should win now? No thanks.
I know no one system is ever going to appeal to everyone, but that one really turns me off.

Yeah that isn't right.

NAF rankings to give an opening/first round seeding is a good idea and I am really looking forward to seeing how that plays out MonkeyBowl this coming weekend. But I don't like the idea of using them as a tiebreaker.

Actually this is not happening at MonkeyBowl. Draw is random and swiss (as normal). What they do is: Before each match the two players compare their handicap scores and calculate the difference. The player with the higher handicap score uses this difference to derive an Inducement Allowance.

straume wrote:Actually this is not happening at MonkeyBowl. Draw is random and swiss (as normal). What they do is: Before each match the two players compare their handicap scores and calculate the difference. The player with the higher handicap score uses this difference to derive an Inducement Allowance.

Ah, my bad. I must of been thinking of another discussion about NAF rank seeding and got the two confused.

But it is using the rankings for something other than just numbers to look at

As straume says, that is not what is happening. [Edit- ninja'd ]

They used to use NAF ranking first round pairings at Pearlies (before I was doing it) and it just meant that everyone used a new 150CR race or one that has an artificially low ranking. For example, Ironjaw (ex-England coach) has a 98CR Lizardman team sitting waiting for the day he can go to a tournament and scalp some big fish.

Pairing coaches on global NAF ranking is distasteful for me. It lessens the chances of those people meeting in the final. It also in small tournaments vastly distorts Swiss (as seen at Eurobowl Denmark) as by the later rounds all the top people are more like to have already played each other, and some weird matches are thrown up. Also, some people would like the opportunity to play Jimjimany or whoever, and admit themselves the random round 1 draw is their best shot of doing so. Why deny them this? Finally, it is a sad state of affairs when we want to 'punish' coaches for being good.

Punish? Don't you go to tournaments to play the six highest quality opponents you can?

I do (hence my increase in Euro tournaments if you look at my record over the last few years). I don't go to tournaments to play the same 1-2 people R1 every single time. Nor would I go to a tournament like this, when clearly the purpose it not to give me a level game, but to 'knock out' at least one of the good coaches as early as possible.

As I thought. I suppose it depends on the size of the tournament? At NAFC would you rather play a new player or someone amazing in round 1, if you hadn't played either before?

At the end of the tournament I would much rather have played the amazing person and had the pleasure of that game. Of course, I wouldn't be human if there wasn't a part of me that at the time thought "ugh, what are the chances of drawing that guy?!", but I only play BB for the tough games now. If the NAFC was just me and 159 new players, I wouldn't go.

lunchmoney wrote:

Greshvakk wrote:

Nippy Longskar wrote:What about SoS based on sum of opponents' NAF win ratios?

I wondered about something like this. Strength of schedule based on a single tournament's results is really saying 'strength based on how people performed this weekend'. Using NAF rankings or win percentages could provide a more 'absolute' base. I mean if those rankings can't on average tell us who is better then why do we maintain them?

As an add on question would you use a coaches overall ranking/win percent or the numbers for the race they were using that weekend?

I massively improve my game and pull it out of the bag and get 510 at a weekend. Only one other coach gets the same result. But because he's played better opponents in the past he should win now? No thanks.
I know no one system is ever going to appeal to everyone, but that one really turns me off.

@ Greshvakk - yep that was my thinking. This is a metric to assess how hard your games were, and I'd base it on overall, all-race ratio as a go-to average. It's still very fallible, beating Joe (poor old Joe) with his Goblins is as valuable as beating him with his Wood Elves, for example. Is beating a strong player who's using a weak team more valuable than a weaker player with a stronger team? I guess that's part of the question.

@ lunchmoney - I don't understand your point about better opponents in the past? In the scenario you describe a tie breaker is required, and the one that gives the least bias towards luck over the weekend and the most towards having played more difficult opposition to achieve the same result is surely the one based on the greatest body of data (overall NAF W/L ratio of opponents)? Or I may be misunderstanding you.