If the red zone is the most important real estate in football then the goal line might be the Holy Grail. From 2002-2008 there have been approximately 8,000 plays in this sacred space (note all 2002-2008 stats fromAdvanced NFL stats and 2009 stats from my own downloads). Of this amount approximately 4,400, or 600 per year, are given to running backs to attempt a TD. Most of these attempts go to “goal line backs” because as some fans lead you to believe these backs will score nearly every time. But how much better are goal line backs in reality when compared to non-goal line backs?
To answer this question I started by determining how often a running back should be expected score from a specific yard line, similar to an expected points calculation. I limited the study to all carries within the 5 yard line since these are typically considered a “goal line carry”. From 2002-2008 there were approximately 4,400 carries that met this criteria with the following conversion rates.
Yard line Conversion rate
1 56%
2 42%
3 32%
4 26%
5 20%
Using these rates you can convert every play into an Expected TD added (ETDA) by comparing the actual results to the expected results. For example, if a running back scored a TD from the 1 yard line they will be given credit for 1 TD and charged .57 TD’s since a league average running back would have scored 57% of the time thus giving them an ETDA of .43. Positive ETDA is good and negative is bad. Additionally, we can divide the actual TD’s by the expected TD’s to determine an ETDA%, which in this case would be 175% (1/.57).
After determining the ETDA for all plays in this time frame I decided to split the running backs by “goal line backs”, which I designated as the running back on each team with the most goal line carries, and “non goal line backs”. By these designations it turns out that the goal line players scored 1,243 TD’s vs. an expectation 1,209 (103%)where as the non-goal line running backs scored only 464 TD’s vs an expectation of 500 (93%), we can consider this a rough “replacement level” value. Based on these numbers it seems that running backs that are trusted to score the most goal line carries score do actually score at a slightly higher rate than non goal line running backs.
Note, how small the difference is though. The average team is expected to get between 13-25 goal line carries per year with the average right at 20 and if the distribution between 1-5 yards is uniform throughout the league they would be expected to score approximately 7.8 TD’s on those 20 carries (based on a 100% ETDA%). If a team used only running backs that made up the non-goal line running backs group they would only be expected to score 7.3 TD’s. If we compared this to the expected TD’s a “goal line back” would score (which is 8.0) you can see that a “goal line back” only scores 0.7 TD’s than a “replacement level back” for the entire season. Furthermore, given that a good portion of these runs are on a first down and the offense has more chances to score means that even less than 0.7 TD’s are lost from a team without a goal line back.
Now most will argue well “wait I only mean the good goal line backs score TD’s all the time”. So I decided to research this a little further by splitting the running backs into two fairly equal groups. All running backs with over 15 goal line carries in a given year and all running backs with less than 15 goal line carries in a specific year. By splitting the groups like this the high volume running backs scored 736 TD’s when they would have been expected to score 701 (105%). On the flip side the low volume running backs scored 971 vs an expectation of 1,008 (96%). Using the same calculations as above a team of all of the league average “best” goal line backs would score 8.2 TD’s per season where as the non-goal line backs would score 7.5 for again a difference of 0.7 TD’s, or 0.9TD’s over a “replacement level”.
No, what I what I meant by the good goalline backs is really the ELITE goal line backs score the most. So what if we limit our sample to only the running backs who were given the most goal line chances over the entire 7 year period? No small sample sizes and lots of attempts time over a long time. Were they able to score at a much higher than average rate? For this sample I looked at the top 5% of running backs in terms of total goal line carries for the entire period of 2002-2008. In this sample the running backs scored 687 TD’s vs an expected 630 for a 109% rate, which while better than the other averages is still not going to set the world on fire. This would equate to approximately 8.5 TD’s for the season, 1.2 more than our imaginary replacement level.
That said there were 3 running backs in this group that really stood out from the rest as shown below.
Player Team Expected TDs Actual TDs Inside 5 carries ETDA %
L.Tomlinson SD 56 73 147 130%
S.Alexander SEA 49 66 113 135%
P.Holmes KC 28 45 76 161%
Tomlinson, Alexander, and Holmes are clearly 3 of the best running backs in the NFL over the last decade but would you think of them as the best goal line backs as well? Maybe the best goal line back is simply the best player or maybe the player behind a line that can block for a short gain?
So how can this help me for the upcoming fantasy season? Since it is fairly clear that if a player is given goal line carries he is probably going to score at a rate close to league average (remember the best in the league only score about 1TD more than the worst) we can look at the players last year who scored significantly more or less than they would have been expected to score and chalk some of those points up to “luck” or “small sample size” however, you want to define it.
Here were the most “unlucky” – ie players that I would expect to score at a better rate this year and may be undervalued in terms of TD’s. Remember that they must still be given as many opportunities to score to really be undervalued. Alternatively, if they cash in more often their passing game and/or FG kickers might be overvalued.
Player Team Expected Actual Difference
M.Forte CHI 6.8 1 5.8
J.Snelling ATL 6.7 3 3.7
B.Jacobs NYG 6.9 4 2.9
L.Tomlinson SDG 11.7 9 2.7
J.Stewart CAR 5.6 3 2.6
Assuming that Snelling and Tomlinson don’t get the same chances they had last year that leaves Forte, Jacobs, and Stewart as the most likely benefactors for an uptick in rushing TD’s. If Forte in particular can continue to get the opportunities that he had last year and start to punch a few in that will be a very significant swing in his value. Some others that are high on the list are Steven Jackson, D. Williams, Chester Taylor, and Maurice Jones Drew. Seeing players like Jones-Drew and Brandon Jacobs on the list of biggest underachievers is pretty surprising but shows even more so how hard it is to judge a goal line back on one season of stats.
What about the flip side. Players who got “lucky” due to small sample sizes last year and might be overvalued this year.
Player Team Expected Actual Difference
W.McGahee BAL 4.4 7 2.6
R.Grant GB 5.6 8 2.4
R.Mendenhall PIT 3.8 6 2.2
A.Bradshaw NYG 2.0 4 2.0
M.Turner ATL 3.2 5 1.8
A.Peterson MIN 12.3 14 1.7
McGahee is obviously the famous one as he started the season out at a blistering pace scoring left and right at the beginning of the year. But, what about Bradshaw on the top of the list as an overperformer and Jacobs on the bottom of the list as an underperformer? Who would have guessed that? Also Mendenhall and Turner overachieved on partial workloads, if you extrapolate out their TD rate from last year into more carries you may overestimate their TD’s even more than by the amounts above. BTW how much is 2.4 TD’s worth for Ryan Grant, well if you look at his VBD of about 61 from last season in a standard league 2.4 TD’s represents about 23% of that value!
A special note should also be made for Adrian Peterson who had 30 goal line opportunities which is the second highest amount since 2002 to only Tomlinson’s 2004 season. He also scored at a slightly higher rate than expected of 114%. If his goal line opportunities go down to even the same level as the 3rd highest total in the league (20) and his rate fluctuates down to 100% he will only be expected to score around 8 TD’s which would be a drop of about 4TD’s. When you get to the price range that you are paying for a player like Peterson that is the difference in probably $20 dollars in an auction for stats that he may need to be fairly lucky in order to achieve again. BTW, Peterson’s career rate is 124% so maybe 100% is a bit pessimistic but wouldn’t be outside the realm of possibility.
So how much is a goal line back worth? Somewhere between about 0.5-1 TD for a real life team but given how likely any RB is to score from the goal line a ton for a fantasy owner. But even more important is to know how lucky or unlucky a goal line back was last year to help figure out their real value this year so you don’t over pay for someone like Mendenhall and help you find a potential deal in someone like Forte - given he doesn't lose his goal line job.
Sorry for the post getting a little longer than I anticipated but I thought it was information that some might find interesting.
To answer this question I started by determining how often a running back should be expected score from a specific yard line, similar to an expected points calculation. I limited the study to all carries within the 5 yard line since these are typically considered a “goal line carry”. From 2002-2008 there were approximately 4,400 carries that met this criteria with the following conversion rates.
Yard line Conversion rate
1 56%
2 42%
3 32%
4 26%
5 20%
Using these rates you can convert every play into an Expected TD added (ETDA) by comparing the actual results to the expected results. For example, if a running back scored a TD from the 1 yard line they will be given credit for 1 TD and charged .57 TD’s since a league average running back would have scored 57% of the time thus giving them an ETDA of .43. Positive ETDA is good and negative is bad. Additionally, we can divide the actual TD’s by the expected TD’s to determine an ETDA%, which in this case would be 175% (1/.57).
After determining the ETDA for all plays in this time frame I decided to split the running backs by “goal line backs”, which I designated as the running back on each team with the most goal line carries, and “non goal line backs”. By these designations it turns out that the goal line players scored 1,243 TD’s vs. an expectation 1,209 (103%)where as the non-goal line running backs scored only 464 TD’s vs an expectation of 500 (93%), we can consider this a rough “replacement level” value. Based on these numbers it seems that running backs that are trusted to score the most goal line carries score do actually score at a slightly higher rate than non goal line running backs.
Note, how small the difference is though. The average team is expected to get between 13-25 goal line carries per year with the average right at 20 and if the distribution between 1-5 yards is uniform throughout the league they would be expected to score approximately 7.8 TD’s on those 20 carries (based on a 100% ETDA%). If a team used only running backs that made up the non-goal line running backs group they would only be expected to score 7.3 TD’s. If we compared this to the expected TD’s a “goal line back” would score (which is 8.0) you can see that a “goal line back” only scores 0.7 TD’s than a “replacement level back” for the entire season. Furthermore, given that a good portion of these runs are on a first down and the offense has more chances to score means that even less than 0.7 TD’s are lost from a team without a goal line back.
Now most will argue well “wait I only mean the good goal line backs score TD’s all the time”. So I decided to research this a little further by splitting the running backs into two fairly equal groups. All running backs with over 15 goal line carries in a given year and all running backs with less than 15 goal line carries in a specific year. By splitting the groups like this the high volume running backs scored 736 TD’s when they would have been expected to score 701 (105%). On the flip side the low volume running backs scored 971 vs an expectation of 1,008 (96%). Using the same calculations as above a team of all of the league average “best” goal line backs would score 8.2 TD’s per season where as the non-goal line backs would score 7.5 for again a difference of 0.7 TD’s, or 0.9TD’s over a “replacement level”.
No, what I what I meant by the good goalline backs is really the ELITE goal line backs score the most. So what if we limit our sample to only the running backs who were given the most goal line chances over the entire 7 year period? No small sample sizes and lots of attempts time over a long time. Were they able to score at a much higher than average rate? For this sample I looked at the top 5% of running backs in terms of total goal line carries for the entire period of 2002-2008. In this sample the running backs scored 687 TD’s vs an expected 630 for a 109% rate, which while better than the other averages is still not going to set the world on fire. This would equate to approximately 8.5 TD’s for the season, 1.2 more than our imaginary replacement level.
That said there were 3 running backs in this group that really stood out from the rest as shown below.
Player Team Expected TDs Actual TDs Inside 5 carries ETDA %
L.Tomlinson SD 56 73 147 130%
S.Alexander SEA 49 66 113 135%
P.Holmes KC 28 45 76 161%
Tomlinson, Alexander, and Holmes are clearly 3 of the best running backs in the NFL over the last decade but would you think of them as the best goal line backs as well? Maybe the best goal line back is simply the best player or maybe the player behind a line that can block for a short gain?
So how can this help me for the upcoming fantasy season? Since it is fairly clear that if a player is given goal line carries he is probably going to score at a rate close to league average (remember the best in the league only score about 1TD more than the worst) we can look at the players last year who scored significantly more or less than they would have been expected to score and chalk some of those points up to “luck” or “small sample size” however, you want to define it.
Here were the most “unlucky” – ie players that I would expect to score at a better rate this year and may be undervalued in terms of TD’s. Remember that they must still be given as many opportunities to score to really be undervalued. Alternatively, if they cash in more often their passing game and/or FG kickers might be overvalued.
Player Team Expected Actual Difference
M.Forte CHI 6.8 1 5.8
J.Snelling ATL 6.7 3 3.7
B.Jacobs NYG 6.9 4 2.9
L.Tomlinson SDG 11.7 9 2.7
J.Stewart CAR 5.6 3 2.6
Assuming that Snelling and Tomlinson don’t get the same chances they had last year that leaves Forte, Jacobs, and Stewart as the most likely benefactors for an uptick in rushing TD’s. If Forte in particular can continue to get the opportunities that he had last year and start to punch a few in that will be a very significant swing in his value. Some others that are high on the list are Steven Jackson, D. Williams, Chester Taylor, and Maurice Jones Drew. Seeing players like Jones-Drew and Brandon Jacobs on the list of biggest underachievers is pretty surprising but shows even more so how hard it is to judge a goal line back on one season of stats.
What about the flip side. Players who got “lucky” due to small sample sizes last year and might be overvalued this year.
Player Team Expected Actual Difference
W.McGahee BAL 4.4 7 2.6
R.Grant GB 5.6 8 2.4
R.Mendenhall PIT 3.8 6 2.2
A.Bradshaw NYG 2.0 4 2.0
M.Turner ATL 3.2 5 1.8
A.Peterson MIN 12.3 14 1.7
McGahee is obviously the famous one as he started the season out at a blistering pace scoring left and right at the beginning of the year. But, what about Bradshaw on the top of the list as an overperformer and Jacobs on the bottom of the list as an underperformer? Who would have guessed that? Also Mendenhall and Turner overachieved on partial workloads, if you extrapolate out their TD rate from last year into more carries you may overestimate their TD’s even more than by the amounts above. BTW how much is 2.4 TD’s worth for Ryan Grant, well if you look at his VBD of about 61 from last season in a standard league 2.4 TD’s represents about 23% of that value!
A special note should also be made for Adrian Peterson who had 30 goal line opportunities which is the second highest amount since 2002 to only Tomlinson’s 2004 season. He also scored at a slightly higher rate than expected of 114%. If his goal line opportunities go down to even the same level as the 3rd highest total in the league (20) and his rate fluctuates down to 100% he will only be expected to score around 8 TD’s which would be a drop of about 4TD’s. When you get to the price range that you are paying for a player like Peterson that is the difference in probably $20 dollars in an auction for stats that he may need to be fairly lucky in order to achieve again. BTW, Peterson’s career rate is 124% so maybe 100% is a bit pessimistic but wouldn’t be outside the realm of possibility.
So how much is a goal line back worth? Somewhere between about 0.5-1 TD for a real life team but given how likely any RB is to score from the goal line a ton for a fantasy owner. But even more important is to know how lucky or unlucky a goal line back was last year to help figure out their real value this year so you don’t over pay for someone like Mendenhall and help you find a potential deal in someone like Forte - given he doesn't lose his goal line job.
Sorry for the post getting a little longer than I anticipated but I thought it was information that some might find interesting.
