|
You drafted two star runningbacks in the first two rounds. You don’t need to worry about your starting backs for the rest of the draft, as you’ve projected them both to have very good seasons, being proven veterans coming off good years. Week one comes around and your two runningbacks each score six points. You aren’t very happy, but it’s week one—not a big deal, right? Next week, both once again fail to reach the end zone, and have put up consecutive underwhelming weeks. Now your opinions have done a 180: your once formidable RB tandem now appears weak, and you wonder: is this going to be a down year for them? The passing attack can’t keep anyone honest? The offensive line just isn’t playing well? How do you know when a player is in line for a bad year, such as Eddie George after scoring seven points in week one last year, or it’s just their bad game(or games) of the year, such as Jerome Bettis and his three point week one? This study will look to see at what point one should give up on a player, and try to pin-point a specific week as to how long a player can be just “starting off slow.” Data informationObviously, looking to find the point when good RBs are having bad seasons is extremely difficult. A significant amount of variables enter this equation, and it is nearly impossible to isolate all the factors. Here is what I did:
The following chart looks at the RBs who started off the season by scoring 20% less fantasy points per game than the previous season. “Started off” can range anywhere from one week through week eight. As you would expect, a fairly high number of RBs had slow starts after week one, since there’s such a small sample. Furthermore, the number of RBs having poor seasons declined in all but one week. It’s very surprising that a full twenty five percent of the RBs in this study were doing poorly by their standards at the halfway mark. The second row shows how many of those RBs with bad starts were RBs who had worse seasons than the previous year, and the third row shows how many of those RBs with bad starts had better seasons than the previous year. “Percentage” shows what percentage of the RBs scoring at least 20% less fantasy points per game at any given week, were RBs with bad seasons. While it’s certainly not a big cause for alarm, I was surprised that 61% of the RBs with a poor opening week went on to have a worse season.
The Amazing Faulk While Marshall Faulk has done enough amazing things to warrant his own article, here is something you probably never knew. Marshall Faulk was the only RB in this study to be at 20% less FP/G than the previous year, at the midway point in the season and STILL finish with a better season. This occurred in 1997. Faulk scored 11.1 FP/G in 1996, in his worst season as a pro. Still, it looked like he was about to post another sub par year: midway through 1997, Faulk had only 404 rushing yards, and just two TDs. In the second half, however, he ran for 650 yards and had six TDs. He scored just seventy fantasy points in the first half, but was still able to post a 200 point fantasy season. The following chart looks at the RBs who started off the season scoring 20% more fantasy points per game than the previous season. The second row shows how many of those RBs with good starts were RBs who had better seasons than the previous year, and the third row shows how many of those RBs with good starts had worse seasons than the previous year. “Percentage”, shows what percentage of the RBs scoring at least 20% more fantasy points per game at any given week, were RBs with improved seasons.
It’s pretty clear that most of the RBs off to good starts after two weeks end up posting even better numbers than the year before. The fast starts by Ahman Green and Corey Dillon propelled them to their best seasons to date. With other RBs, history hasn’t been as clear-cut. However, each week that the RB continues his bad start increases the chances that he’s going to do worse this year by about five-to-ten percent. Only four of the twenty-two RBs were able to rebound from bad starts (through four games) and post better numbers the following year: Corey Dillon and James Stewart in 2000, Marshall Faulk in 1999 and Karim Abdul Jabbar in 1997. Dillon’s 278 yard rushing game with two scores against Denver, and James Stewart’s 116 yard, three score performance against Tampa Bay were big reasons they were able to make up for the slow start. Jabbar had TWO three score games the rest of the way, and we all know about Marshall. In fact, after those four weeks he averaged 200 combined yards the next three games. However, this isn’t the end of the study. Look at this hypothetical RB
If this RB had averaged 11 FP/G the previous year, he would have been very disappointing with just twenty fantasy points through four games. Sure enough, he had a down year. But, in the fantasy world there’s nothing you can do about the past—you have to make your decision on the future. After four weeks, you probably wouldn’t treat him like an 11FP/G RB. In fact, he wasn’t: The rest of the year he was a 12 FP/G RB! The question now is how many of the RBs listed above with “bad” seasons had bad seasons, or just bad starts that made their season look bad. In fact, there is a real life example of this. In 2000, Ahman Green averaged 15.7 FP/G. Last year, he averaged 16.7 FP/G. However, after his opening day 30 against the Lions, he averaged 15.6 the rest of the year. When making decisions on Green, the thirty he scored in week one won’t affect how he does the rest of the year (he actually did ever so slightly worse), but it could alter his year-end totals. So how do we resolve this? I looked at how many FP/G each player scored the rest of the year, after each of the first seven weeks. I then compared it to their FP/G from the previous year to see how many of the guys with real slow starts (20% or less FP/G through the first X amount of weeks) actually played worse in year N+1 the rest of the way. As an added bonus to you, the fantasy player, I looked to see how much worse (or better) they did the rest of the year, compared to year N-1. RBs with BAD Starts (20% or less FP/G through the first X weeks)
FP% shows how much, the average player who improved (or declined) their FP/G from a year ago improved (or declined) for the rest of the year. Of the twenty-two RBs off to poor starts after a quarter of the season, seventeen of them had less FP/G than they did in the previous year FOR THE REMAINDER of the season. In fact, those seventeen players saw their FP/G drop by an average of 23%. For example, Curtis Martin finished fifth in fantasy points last year, scoring an average of 14.9 FP/G. If he got off to a slow start, scoring 11.00 FP/G through the first four weeks, one would expect ON AVERAGE the following to occur: He would have a 77% chance of still scoring less than 14.9 FP/G FOR THE REST OF THE YEAR, and would be likely to average around 11.5 for the rest of the year. As expected, the trend to show a decline is less when looking at the rest of the season compared to the whole season. Twenty of twenty RBs with poor starts through week seven had worse seasons, but twenty percent of those RBs scored more FP/G the rest of the year than they did the year before. Another note: For the RBs that did improve the rest of the year, at just about every stage their improvement declined. While the RBs who improve the rest of the way after a bad week one improved by an average of 22%, those with poor starts through seven weeks who improved, did so by only 8%. The “bump” in week three can be almost entirely attributed to the amazing Fred Taylor. Through three weeks he was awful, and there were only eight RBs left that still improved after a start like that. Taylor was dynamite for the second half of 2000, joining Marshall Faulk in the rarefied air of 26 FP/G. How about the guys with good starts? How were they affected when we looked at how they did the rest of the way, as well as by what percentage they improved or declined?
Through the first two weeks there’s not much pattern, but players with very good starts (20% or more FP/G than the previous year) seemed to continue to have better years than previous for the rest of the season, about 70% of the time. Week six was a little odd, but three guys who were right on the “fringe” of improving or declining, all moved from improved to declined from week six to seven. All four scored within six percent more than their FP/G for the rest of the year after week six, but had big enough weeks to move into the declined column for the rest of the year after week six. This is why both the % Imp is curiously low, and the average % of fantasy improvement is so high (the loss of the players that watered down the average). SUMMARY What can we take from this? I didn’t expect to see such a strong correlation between weeks. One of my fantasy trading problems is I consistently trade early season stars for last year’s stars, expecting the law of averages to work for me. Based on history, this doesn’t appear to be very sound (and is likely why I’ve had more than a few trades go bad for me). While it’s important to remember we’re not comparing two different guys (such as a RB 23 in preseason who is RB5 after three weeks, to a RB5 in preseason who is now RB10), it does appear that it doesn’t take too long to see which guys will improve from past seasons. Four weeks worth may not seem like much, but the fact that 77% of the time the RBs still performed below average the rest of the year is enough for me to say week four is a rough estimate of when to “give up” on a player. How does this help you in a fantasy league? You may want to see if you can get “preseason” value for your stud after three or four weeks, if he’s not performing up to par. Another interesting thing is that the trends are stronger in the RBs that start slow. The fast starters don’t tend to hold their value as strongly, and I can venture a couple of guesses as to why. One, as Doug Drinen has showed before all good players decline (on average). Two, it is to be easier to be bad than good. That may not be very scientific, but true in just about every facet of life. Another important factor to remember is that not only do the chances of your stud improving his play go down week after bad week, but his upside gets more limited as well. You’re probably hoping for last year’s production, with not much room to improve. So how do you go about acquiring the guys on their way up, and dropping the dead weight? One possible reason this study showed a lot of early season results holding strong throughout the year is that a coach can gain confidence (and lose it) very quickly. This is especially true if the team is winning games. Last year not only did Eddie George start off slowly, but the Titans lost their first three games. Fisher went to a more vertical passing attack, and the Titans started winning. Fisher lost faith in Eddie, and fantasy owners should have as well. On the other hand, Curtis Martin had his best season of his career last year. Through five games the Jets were 3-2, but Martin was averaging 120 rushing yards and 1.3 TDs in Jets wins. Coach Edwards stuck with what was working, and Curtis had his best season. While fantasy players often get caught in a vacuum and just try and isolate the individual player, how they affect their team is a very important aspect to watch. When looking for this year’s studs and duds, I would look at the following things: When the team wins, is he a big part of the victory? What’s his yards per carry looking like? Is the player playing very well, and impressing the coach? Lastly, are his fantasy stats coming from categories you can count on (like yards), or are more fluky things affecting his fantasy point totals (touchdowns, fumbles). To follow up: Each week I will be looking at guys who surprise and bust each week at not only RB, but QB and WR. We’ll get to look together, at what guys you should be going after and who are the guys you’ll want to put on a competitor’s roster. Exclusive to Footballguys.com |
