In the discord chat last week I mentioned that there was a big disparity in points scored for the kicker and dst positions. After 5 weeks the highest point scorer from these position was Tittsburgh Feelers with 131 points, the lowest The Broken Soldiers with 44. That’s a 87 point disparity or 17.4 points per week, quite a big deficit to make up from the remaining 7 positions.
This led the discussion to talk about the merits of having kickers, with a few people suggesting not having the position at all. For the record I don’t have an issue with kickers, I don’t see how they are more than random than touchdown scorers.
However this got me to thinking how influential the different positions have been. To look into this I have looked at the results of the 40 games played so far. Then I have seen if the result changes if we take away the points from a position group. (We have to remember that for RB and WR there are two players in this position group so you would expect these positions to influence more games). The results are as follows:
This shows the kickers has influenced the results of 3 out of the 40 games so far. This doesn’t sound like very many until you look at the other positions. This is consistent with the TE number and only 1 less than the QB number. Effectively the kicker has been as influential as the QB in our fantasy league. In fact if we had removed the RB position then 35 of the 40 results would still have been the same, is this a reflection of the lack of every down backs suppressing RB scoring?
All of these numbers seem quite low to me, and that maybe a function of how the Originals league is panning out this year. I don’t have the numbers to back it up, but it feels to me that this year there seems to be a definite spread between the good and bad teams with a few in the middle, previous years it felt a lot closer at this stage.
Good work very interesting analysis.
With regard to QB it backs up the theory of taking one late or streaming them.
Also backs up the zero RB theory as illustrates the value of WR’s
Also shows the importance of having a good DEF for that particular week.
Interesting figures there Chris - if I’m reading them right it also shows how decisive Defences have been (couple of big games for the Vikings and Chiefs would definitely have affected results in many leagues) and surprisingly also the flex spot which has affected more than QB even.
When I have a bit of time later in the season I may try to do the same for Sean’s league. His flex spot allows TEs too, will that make any difference? And yes, I agree the league has felt a bit more spread out… Don’t think we had two 5-0 sides last year or 2014? And I wonder if the average winning margin in games is up a little this year compared to previous
Love the analysis
My view from what you have said is I would think it makes sense to simply give three points for all kicks. Don’t bother with five points for 50 plus kicks it makes zero sense. Also don’t take points away for misses.
I think this is a sensible compromise.
I think defences are a stable in fantasy and I’m leaning more towards a system that rewards only sacks,turnovers and tds with zero focus on points conceded against it. The reason is this balances out defences better again in my view.
I’m not taking them away and I’m not reducing the good ones I’m just balancing out the impact.
You highlight something important in your analysis how the rbs impact is down and in Jons league he has an option to only use one but he has added flex positions to use a second if you like.
Personally I love that idea and is something that will be discussed out season in detail.
Interesting but I’m not sure if this proves anything or if we are drawing the right conclusions here -so if I understand these are the changes in the results if we move one position out of the games played so far?
Therefore aren’t the results then dependant on the remaining population not just the removed population? So if a game had a 40 game lead without kickers the kickers would be irrelevant whilst if it was a 1 point lead virtually every kicker would have made a difference?
Also as I think this way was worked backwards the actual kicker chosen makes a difference so in some cases you removing the max that week and in some cases zero probably
I do think your on to something having said this maybe a MAX analysis on a spread analysis as your have mentioned previously
Also re RB WR a low hanging fruit analysis would be to see how many RBS or WR are used in the flex spot
Completely agree this doesn’t prove anything, I just thought it was one interesting way to look at the data we have. And this is only up to the end of week 5 so 40 games, will be interesting to review at the end of the year.
Yeah, however this does show many games were influenced by the kicker. So like you say the kickers will only influence the games that are relatively close. However the chatter in discord was that peoples impression were that kickers were too influential. This shows by removing them only 3 results would change so they are not having that much impact. In fact the only people that can feel hard done by the kicker position are Chris (now not Feelers), Noel and Al (Feelers).
As Jon eludes to we both feel the average winning margin this year feels higher than last year and this may mean that kickers are less influential. Might be one to look into next week.
Another point is that kickers are probably more influential in a 16 team league than a 10 team league. In the 16 team league the average score should be lower as you get into 3rd choice WR and 3rd down backs being played regularly. However as the kicker points are more random the average points from a kicker are probably not deflated in a 16 team league (again I have nothing to back this up). So maybe there is some rationale to reduce the points in a 16 teamer.
Interestingly Al’s mega Vikings defence haven’t won any games for him. His team is so stacked he’s winning by enough points that the Viking’s heroics haven’t had any impact. In fact the Vikings defence have cost him a win, in week 4s three point loss to Dean the medicore 8 points scored by the Vikings was beaten by the 15 points accumulated by Denver which was enough for Dean to take the W and the bragging rights.
Obviously as Shani points out if they had been used in a close game it would have had an influence, but I’m only in one league this year so have a very small sample of data.
The talent pool at RB is smaller so for me they are more valuable in the earlier rounds (coupled with the fact everyone seemed intent on zero RB so they were being undervalued). I used to do zero RB when the advice from all the analysts was to target RBs early, that created better value at WR. I’d have happily taken a WR in those first 3 rounds if they were top of my board. I have an equal number roughly of RBs & WRs. The later round RBs tend to be backups whereas later round WRs still tend to be starters and can be relatively productive. I was able to pick up people like Larry Fitzgerald, Travis Benjamin, Allen Hurns & Philip Dorsett round 4 onwards who have given good production (16 team leagues).
To be honest, in our league there seems little correlation between draft strategy and those who have performed well (I’ve not done deep analysis). There are examples of good teams with zero RB, zero WR and those who took a mix. I’ve been able to improve my team significantly through trading which has maybe been as important but core of my team is still a strong RB paring.
I know Sean has done this for previous years eg Tracking who started their draft Wr Wr or Rb Rb or one of each, and seeing how many wins each got and who reached the playoffs. Was interesting stuff and maybe good to see again when we have this year to add to the stats
Trust me I’m already building a deep analysis and the results will surprise people even thou we are only five weeks in.
I won’t release data till it’s all said and done…
But it’s very interesting and may surprise many people
I keep a tracker analysis as my ‘strategy’ is not based on positions but on something else. I agree some of the positional strategies used have enabled to get value in a counter strategies. I think the drafting in the TFP leagues has reached a mature no stones unturned stage and with it there has been a lot of taking sleeper type players in advance of probably where you’d like because you feel someone else is going to pick that player soon and that’s true maybe more so than ever in the challengers league
I maybe wrong but injuries and missed games feels like it’s on the rise and I found it interesting there was a general ‘feel’ that some teams were banking on back up players coming in which is obviously how to find value later in the draft but sacrifices were clearly made strategically early by those teams willing to do that and it’s providing value early
Will check on this but I feel the feature back is now coming more back into play giving more touches to starting RBs than the RBBC of recent past
the last thing I noticed a s the difference in activity between originals and challengers league - not sure exactly why…
So I’ve updated some analysis to the end of our fantasy regular season (week 13). With all results coming from just the data in our league the impact of positions is:
This is has been calculated by looking at the final results of all 104 regular season games. Then recalculating the result with each positions score removed. So for example if the same teams played but without the QB then the result would have changed in 11 out of the 104 games.
Shani bought up a good point before that one of the limitations of this is that this is biased in regards of how good the rest of the team is. To try and address that I have created a schedule of 10,000 games with teams made up of randomly selecting players score from Week 1-13. The sample size and random sampling of the teams should remove the underlying quality of the remaining teams.
So for example the match up may look something like this:
(I made this up so may include players when on bye or if they weren’t picked but in reality it will only pick players that were on the active roster in the originals.)
Running the same analysis presents the below results in terms of positional impact. Which are probably a bit more accurate. These are still biased by who was played in the league and what position they were played.
Using the same sample of 10,000 games I also calculated a players impact. They scored a +1 if the when their position was removed their team went from winning to losing, and a -1 when their team went from losing to winning. Using this the top 25 players were:
Obviously nothing groundbreaking in who is in here, but maybe some interesting nuggets can be taken in terms of where the positional groupings are. I wouldn’t expect to see two defenses so high for example.
Also interesting to see Ryan as best QB despite Rodgers and Brees having more points. With Rodgers it’s probably a toss up, but Brees is quite a few spots behind. This is probably because Brees had more really good and pretty bad weeks, but Ryan was just consistently good so picked up less points but didn’t lose you any games.
Finally to wrap up the worst players in the TFP Originals from my dubious method:
These were players who were consistently picked (in my case I played Duke Johnson Jr 9 games as I had nobody else at RB) but constantly delivered nothing…
I love this type of analysis.
Also anything which supports my theory that Darren Sproles is totally xxxx gets my vote. He looks amazing for about 2.5 minutes of every game and the commentators can’t wait to heap praise on him.
