Gents,
Whitebread's post about the Gophers mentioned/linked to a bunch of advanced stats, which is so Cracker. "Hey look I read all these advanced stats stuff because I'm smart. I probably can explain only a fraction of these numbers to you, but... HEY LOOK they say the Gophers are good!"
I keed, Whitey. I'm sure you're a regular reader of this stuff and jerk off to IsoPPP in your spare time.
But his post was a good reminder for me to post our own league stats. A brief review, even though everyone should know the drill:
I began doing these about 7-8 years ago - a rough season when I felt like every team I played had their best week against me - to determine when the luck of scheduling was removed, who was the best team.
I started tracking the scores each week to determine how we all fared in comparison with each other. Points Scored and Points Against can give you some idea of this info, but can be skewed by outlier weeks.
 |
| TOTAL POINTS SCORED |
 |
| TOTAL POINTS AGAINST |
There should be nothing surprising about those two tables - they are simply the data. Points Scored and Points Against for each team by week, summed and sorted. That's it.
The only thing above that might be interesting is where the Points/Loser Bonus $ has been allocated thus far. Bill and I collectively have earned 2/3 of the Bonus $ thus far. Surprisingly, Gow hasn't earned any while having one of the stronger teams.
Let's talk Power Rankings...
These are RANKINGS - they RANK us in comparison with each other for each week. It doesn't matter if the top point scorer got 200 points one week, and the following week's top points scorer only got 100 - for the purposes of ranking against the other teams, they both earn a 1 for their respective weeks.
 |
| POWER RANKINGS |
Don't act like you're not impressed! I narrowly own the top spot, just barely ahead of both Dallus and Gow. The three of us are all very close, and the two of them have probably been more consistent than my team.
Those last 3 teams all look pretty bad by these numbers. Whitebread has only had one week where he was higher than 7th (and he lost that week thanks to Russell Wilson's MNF heroics). Crabs started off about as bad as you can, and has played ok enough to win 3 in a row while not looking overly impressive. Juan's team is just plain bad, but as we all are aware, it's 100% injury-driven.
The biggest loser when looking at this is Mr. Fredricks. A pretty darn good team but is currently 1-5. Poor Bill.
As you look at these there is a distinct gap between the Top 5 and Bottom 5. And that becomes even more apparent when you look at the All-Play Records:
 |
| ALL PLAY RECORD |
There isn't even anyone that close to being .500 -- there are 5 good teams and 5 not-so good teams. But Kenni (surprise surprise) and Crabs have lucked their way into a better records, and Bill has just had some of the worst luck.
Let's look at schedules...
 |
| STRENGTH OF SCHEDULE |
Most probably assumed to see Bill at the top, and he's very close (he also has a 2.6pt edge in Total Points Against). But the owner of the hardest schedule-to-date is yours truly. Aside from the first game, I've played a top half team each week including four Top-3 finishers.
Bill has had a tough schedule, but some of his misfortune is just bad timing and dumb luck. When his team has done well, he's played someone slightly better. When he's had easier competition, he's had down weeks.
Whitebread, you may remember, had one of the most difficult schedules ever last year - thus far, the fantasy gods have not shown him the proper regression.
Gow/Dallus both have very good teams this year, but their records are certainly buttressed by playing mediocre-at-best competion thus far.
If you've read everything to this point, it should be no surprise that Kenni and Crabs have had the easiest schedules.
Who have I not mentioned at all yet? Crapswell and Titwarmer. There isn't that much to say about each... This article focuses more on outliers and regression candidates - both of them are more just middle players. Ned's got an above-average team and should probably have a better record, but his schedule has been a little tougher. Andy probably deserves his 2-4 record.
So there you go. First batch of Power Rankings for 2014. What else did you find interesting? Care to refute anything I've said? Let me hear it in the comments.
Nothing to refute really but to garner a discussion, on the last table the median opponent rank is probably just as relevant, if not a better measure, as average opponent rank in measuring difficulty of schedule. For instance, by using average (4.5 vs 4.67), Fredericks has a more difficult schedule but by using median my schedule is more difficult (2.5 vs. 4.5). In looking at them more closely, I'm guaranteed 2 victories by virtue of playing two #10 teams but to win any additional games I would need to be a top 1 or 2 scorer. Fredericks on the other hand has no guaranteed victories but could be .500 or better by scoring 4th in the league every week. If I finished 4th every week I would still 2-4.
ReplyDeletePlease don't confuse the comment as me saying I'm getting screwed. My record is probably exactly what it should be based upon my own performance (I won one I shouldn't and lost one I shouldn't) and Fredericks definitely should have a better record than I do but he doesn't.
:2cents:
Different numbers are going to show different things. Median is flawed too. If you changed both your 10's to 4's, you'd still have the same median, yet I think we'd both argue your schedule was much more difficult. Two guaranteed victories (out of 6 total weeks) is a big deal. Ask Bill, who only has 1...
DeleteMedian and mean can both have merit for discussion. For what I'm trying to evaluate here, I think avg strength of opponent is a better indicator. I think each week should be weighted and taken into account equally.
I will say... one thing I've thought about but never invested that much time into figuring out is determining when teams play better against other teams.
DeleteFor example, the season I started doing this, I had like 4-5 weeks in a row where a team played its best game of the season against me - didn't necessarily make them the top point scorer for that week, but it was their best performance of the year. I specifically remember Kenni having just a godawful team score like 2x as much against me as he did the previous week.
I suppose I'd have to rank the weeks (based on point totals) for each individual team, then determine if there was a correlation to opponent.
Sounds like a lot of work...
Good analysis. The top 3 teams probably all deserve to be 4-2 rather than two of us be 5-1, but that's the luck of fantasy football. Plus it goes to show that consistency is important.
ReplyDelete