Warrior General Games

Owners...please post on the other thread I have set up! Thanks! -GSUFan513

What's the difference between what you're doing and what I'm doing?

Jordan - send me a PM when you want them and Ill do it for the Pigs - GM here and pretty active on boards

I have no idea wtf is going on because my brain is fried from 4 days of recruitment. When you guys sort this out between yourselves, make a statement a 4 year old could understand, because that's my brain capacity right now.

Well as another thread said we could have an OWNERS power rankings, however we should also have a FORMULA ranking. Isnt anyone else in leagues where someone has developed a great formula for ranking teams based on strength of schedule, record, score, ect.?

Ok here's a forumla. It's called the Oberon Mt.

(avg score x 6) + [(high score + low score) x 2] +[ (winning % x 200) x 2]/10




In English:

#1. Multiply the team's average score by six. Average score is the very basic stat to judge a team's prowess.

#2. Add the team's highest score to their lowest score (Deviation), and multiply the result by two. Over and above the average score, the Deviation gives more importance to a team's highest scoring game, while also punishing a team a little more for their lowest score.

#3. Take the owner's winning percentage and multiply by 200, then multiply that by two.** This portion of the Formula more than anything rewards and punishes for all the little intangibles associated with coaching a fantasy team. For instance, an owner that continues to win despite a less-than-impressive lineup is rewarded over and above their lagging average score.

#4. divide the total by 10.

** I know, multiplying by 200, then by two is the same as multiplying by 400, but doing this way shows that multiplying the winning ptc. by 200 first results in a number in the general area of the raw average score and raw deviation. Then multiplying the result by two shows the 20% weight.

The result is the Power Rating . . . or, if you like, the Potential Rating, since it actually is meant to judge the potential score a team might be expected to score on any given weekend compared to its average score. Remember, as the season progresses, a team's average score changes more and more slowly as the number of games included increases. The Power Rating formula takes not only the average score into account, but also recent extreme high or low scores, and winning or losing trends. Obviously, you will need a few weeks of data before a viable PR can be arrived at which is why I suggested starting at week 4.

Definition of Terms:

Average score - Total of a team's points scored divided by the number of games played.

Deviation - Take the team's highest score and add to it the team's lowest score.

Winning Percentage. - Divide the team's number of wins by the number of games played.




Credit:http://www.justlacrosse.com/omffl/powerrating.htm

I copy and pasted and kinda wove my own words in there. What does everyone think? This will be the formula thread and the other can be just us voting, I guess. I'm confused what the difference between what I originally stated and what the other user is trying to do, but whatever I guess.

I don't agree with highest score and lowest score too much. You're going to have some 100 point games, which I'm sure that doesn't account for - although it is multiplied by 20% which helps negate some of it.

We can give it a shot and see what happens. Apply it to another league such as http://goallineblitz.com/game/league.pl?league_id=29 and see what happens.

I mustered up the boredom to do an Excel spreedsheet that calculates this ranking and uses very easy inputs. If someone with hosting capabilities wants to post or PM me their email address, I'll send it to them so we have an example to go off of.

Edit:

I finished the spread sheet and it doesn't tell us much that records won't in the end. The winning percentage plays a very large factor in the equation as it should. It's really only a way to separate teams with even records. Unfortunately, it is not dependent on defense whatsoever, which is a very poor judgment for this kind of game, but is good for fantasy football. I'd highly recommend against using this formula unless defense was somehow taken into consideration whether through points spread instead of average score and max spread/min spread. That would realistically only take another line on each sheet in the spreadsheet, but I'm not willing to go through and re-do that.

Hey Pike if your bored play around with this formula, I think Im going to do the same:

4 variables- and we will have to come up with the weights to determine how important each is:

Strength of Schedule - Sum(opponents winning %+1 to date)/# games played. By week 15 this should come pretty close to 1.5 for all teams (+/- due to the team in questions record not being factored in)

Offensive Differential - Points scored in game-average allowed by opponent. Add up and divide by # of games played.

Defensive Differential - Average scored by opponent - points allowed in game. Add up and divide by # of games played.

Winning percentage

About everyone's going to have the same strength of schedule in our league. We all play each other. It isn't like the NCAA where they have 300+ teams and you play 25 of them. All teams with lower records will have a marginally higher strength of schedule than team with higher wining percentages since teams can't play themselves.

Edit: I just had a thought of if we could somehow combine opponent's win % with the margin of victory as a variable.

Hmm, well for that you can just straight multiply [margin x % as a decimal] and decide how much that variable should be weighed. So this is just off the top of my head so the numbers can be adjusted.

[(average points scored x 1)+(average points allowed x 1)+(margin x opponent winning % as a decimal x 8] / 10

So hypothetical...

Situation 1: Team A wins 20 - 13 over team B
Team A winning % coming into the game as a decimal = .500
Team B winning % coming into the game as a decimal = .750

So team A would get 7.5 points

Situation 2: Team C wins 50 - 3 over team D
Team C % coming into game = .500
Team D % coming into game = .100

So team C would get 8.98 points


So what this really would do is kind of make "quality wins" count significantly and beating up on the bad teams wouldn't really help you much. A blowout against a weaker team gets you only 1.48 more points then a close game against a good team.



IDK, that's my rough idea off the top of my head, the ratios should probably be altered, but that's my idea for now. What does everyone think?

It's a matter of finding appropriate balane for the variables, too. I am not a mathematics major by any means, but have a very good understanding of numbers and football. As my finals wrap up, I'll consider taking a look at it. I wouldn't expect any major breakthroughs from me though - so i you guys can find something that works or generate ideas, I'd be more than happy to analyze them.

Personally, I think it would be easier for each owner to just send in their rankings and comments, sans their own team.

Originally posted by bluepike


Personally, I think it would be easier for each owner to just send in their rankings and comments, sans their own team.

Which is exactly what I suggested at first. I just figured if someone wanted to do the formula, I'd make one up for them. However, I'm perfectly happy doing a owner based vote. So all owners who are interested in sending in power rankings starting week 4, please inform me either by responding in here or by a PM.

We should do owner or designate voting for each team with each owner not including their team. If we do end up doing this, we should stress the importance of certain things such as quality wins and scoring margin. Some people may want to pull directly from the rankings, but a lot more goes into having a solid basis for an opinion/argument.

I'm in though.

Ok so an owner can designate a player from their team to do the rankings or they can do so themselves. Add your name to the list if you would like to join the power ranking. You can send the power rankings to me and I'll figure out the average.

Add your name to the list if interested:

1. bignasty31 - Casablanca Golden Eagles [Elephant]
2. bluepike - Cleveland Bombers [Elephant]