We Measured "Luck" in the NFL. It’s a BIG deal.

one foot not even a foot a few inches changed everything on this play let me show you real quick go ahead and pause it change the angle play it watch it watch it pause it zoom in on that enhance it enhance it one foot the worst punt of the night a halfhazard comet that's going to come crashing down from the clouds only to hit an under deserving lower leg bone a Twist of faith that undoubtedly altered the trajectory of this game but was this luck and better yet how big of a role does Luck play in the NFL in sports so football is far from random if it was random you and I could be out there right now the RAC uh we can't kcks still that doesn't mean that luck can't play some type of role in this dance to understand luck we need to First Define it because if we can Define it then we can begin measuring it luck is a chance event it's a random occurrence where one doesn't maybe have any control over the outcome it also gets misunderstood misused and confused with good in bad fortune something someone would at least have some partial control over the outcome with and luck is also simultaneously this medication for losing fans most don't deny its existence instead the debate in really the doubt starts a little bit deeper with luck so there is a general sentiment that while luck may exist it is evenly distributed I think one hypothetical example perfectly sums this up imagine a head coach explaining a loss due to luck should he choose such a path he may no longer be a head coach at the conclusion of the press conference suggesting that luck plays a role in the outcome of a game immediately violates any sense of responsibility it's Shifting the blame from thyself to thy Universe the question is is this accurate does luck balance out in football to understand virtually anything we kind of need something we can compare it to for example football is big [Applause] in other ways football is small how well football games on average are longer in game length than basketball baseball hockey and that more literal version of football but American football is also smaller in the number of possessions and smaller in terms of games played in a season this is crucial when it comes to luck even for those that kind of get this they maybe don't fully appreciate the scale so here's an example in a typical NBA game there are about 210 total possessions in football each team typically has the ball 11 to 12 times a game we'll round that to 25 total possessions just for a moment imagine the Celtics playing the Pistons in a game with only 25 possessions where two teams play one six-minute game a lesser team will have a much better chance at upsetting a better team with fewer possessions because while the sample size is so small luck only needs to intervene in a handful of possessions to change the outcome roughly speaking the less volume that we have the more variance we can see in a data set the more that luck can play a role and since football games have so few possessions compared to other major sports luck plays a big role in this way of course this isn't the only variable as we compare Sports on the luck Continuum but it is an important one additionally of the four major American Sports football is the only one that doesn't involve a series in the postseason it's one and done so again Advantage luck relative to other sports so how do we measure luck within the game of football I spoke with Tom Bliss who's a data scientist for the NFL and he mentioned they had identified four different actions or essentially chance or luck events the first dropped opponent in interceptions if your opponent just drops an interception you are lucky you should be thankful because well interceptions are important if they just short circuit and get bricks for hands that is a crucial play and similarly when your opponent drops a pass you are lucky and you're especially lucky if they got some green grass in front of them when they do it because well they might have ran for a few more yards thanks to nextg Stats we can see how many yards they probably would have run for and we've waited for this if your opponent just whiffs a kick not close you are lucky oh my good we're not counting blocked kicks here so again this is a rather lucky event and on the other hand if your opponent makes a kick you are slightly unlucky luck swings both ways perhaps most importantly fumble recoveries now some teams are good at generating fumbles that's a form of Good Fortune you have some control over that but the moment that ball comes out it is a chance event it could bounce many different ways of course not all fumbles are created equal we had to account for this for example Tom Brady getting sacked and then fumbling about a 50/50 chance the defense is going to come up with the ball here that is completely different then if the ball gets thrown down the field the receiver takes off gets hit in the secondary and fumbles about a 75% chance the defense gets the ball typically here so we waited for the different types of fumbles now of course there are going to be nuances to this anecdotal counterpoints both ways but but in the NFL these plays are almost entirely chance events still some of you might be itching to recite your fourth grade head coach luck is when preparation meets opportunity this is you hear me out on this we looked in the correlation between season games and even from one play to the next and we found virtually no correlation so if you got lucky on these plays throughout one game three games games over the season or just even a few plays there was nothing to say it would continue that makes these four plays chance events using the NFL's four categories we can examine the effects on win probability a win probability added model quantifies the effects from one event or action to the next as it relates to win probability you may have notice this just tracking a game on your phone for example if a team fumbles in the fourth quarter the percent that change as far as the likelihood they will win the game will drop significantly it might not drop as much if they fumble in the first quarter though we're starting at 2017 because this is when the NFL could start doing player tracking with nextg sports and all the way at the bottom we have the 2021 Tennessee Titans the Titans actually benefited from the Bengals dropping passes in this game however they did not benefit from the fact that the Bengals nailed all four field goals two of which were 50 plus yards remember luck can swing both ways now if Evan mcferson had missed any of these four field goals we would have said well that was lucky for the Titans but he made all four so it was unlucky for the Titans because he's an average NFL kicker on average he should have missed at least one of these field goals and before we start saying these field goal is a result of the Bengals driving into the Titans territory suddenly the Titans did have control over over remember we are isolating specifically for when that ball hits the foot that kicking event and outside of blocking the kick the Titans have no control over which way that ball will fly if it will go in or not it is a chance event think of it this way the more you leave up to luck the more luck will decide your fate that's not an ideal strategy so did luck balance out for the Titans here obviously not they didn't even get a chance for it to balance out because they played one game and well mcferson made all four given our metric that was quite unlucky for the Titans since this was a low-scoring game each one of these field goals carried significant relative weight for the game including that one to win the game is time concluded this is partly why win probability added model isn't perfect and why the Titans were such an outlier football is not random the the Titans had plenty of chances so to speak they obviously didn't lose on luck alone but this small sample size is the reason why we see the greatest parody across the four major American Sports in football so what about the teams that got the luckiest in the post season maybe there's more we can learn here if we look at the top of this chart so if we scroll up we find wait a second enhance it is that that freaking foot again enhance it what contributed well amongst other things that tibia did help in fact the Chiefs recovered four of the five fumbles they forced 80% including this one fumble offensively they also recovered four of the six fumbles they had meaning eight of the 11 fumbles bounced their way the Chiefs also benefited from Jake Moody's missed point after some people confuse this is a blocked kick but if you watch the replay he toe punches it almost directly into line that's it's not a normal angle for a point after bear in mind Moody had set the record for Point Afters just a few weeks prior so this was lucky for the Chiefs maybe you're still in disbelief though maybe that youth football head coach's speech is still rattling around in your head and the idea of luck influencing something as sacred as the game of football that's just sacrilegious uh if this is you then consider a few other points here if if you had a positive luck score in a postseason game you won 69% of the time in fact since we have been able to track this metric three Super Bowl champs have been in the top seven out of the 80 playoff teams I will keep stating this though luck is not everything the 2018 Patriots one of the most unlucky teams and they won the Super Bowl there are anecdotes like this but consider something else I discovered if we look at the top 10 Single playoff luck score course nine of them won their game assuming that luck evens out over the course of the playoffs would be incorrect there just isn't enough time and while there's no correlation from one play to the next amongst these four actions the NFL identified there's a decent correlation between a positive luck score and winning what if we looked at a larger sample size does luck even out across the regular season with more games we can see that luck has evened out considerably the greatest outlier being the 2023 Vikings who experienced about a minus 10% luck per game still this was over 170% cumulative luck for the season looking back I noticed that many vikings fans were very quick to point out this bad luck but downright dismissive and not as boisterous about the incredible luck that they enjoyed just a season prior it is a touchdown on their last I don't think that that is coincidental so since 2018 there have been 94 teams that have a negative cumulative luck score and how many of these teams might have cracked the postseason if luck was more in their favor if they had won maybe one or two more regular season games how much different would history be it's worth noting that there is an error in this data set all the way at the top we have the 2019 Seahawks and the Seahawks have never been lucky they just work harder than everybody else and anybody who disagrees with this shut up throwing out this glitch we see the Chiefs on top again the Chiefs and Seahawks had roughly 200% of cumulative WPA which gives the impression that they maybe won two games off luck over the course of a season but in reality these games are very close they might have only needed 10 to 20% of WPA to win a game so they could have won multiple games over the course of a season due to luck isolating for that Seahawks team I remember that magical ride they had as a fan and I even spoke with Ben Baldwin about it who was a rider for the athletic of the time and covered the Seahawks he had the following to say just 12 games into the season the Seahawks were 8 and one in games decided by a touchdown or less in five of those wins here are some notable luck plays Cincinnati misses is a 45 yd field goal in the third quarter Seahawks win by one rosberger gets injured halfway through the game Seahawks win by two sirine misses a potential game-winning 44 yd field goal Seahawks win in overtime Seahawks win the coin flip in overtime and go on to score a touchdown on the opening drive mlin misses a potential game winning 47 yd field goal in overtime nothing the Seahawks did caused field goals to sail wide right or left or coin flips to come up in their favor but they benefited significantly from these events and went on to win these games they had the smallest Point differential of a 10 and two team ever and their games were historically back and forth again luck was very much on their side so uh this might not be a glitch it appears as though even over a regular season luck can significantly change the trajectory of a team obviously this model isn't perfect there are other forms of such as blown penalties injuries or literal coin flips measuring for these actions and waiting for them right now is a little challenging but I find it difficult to deny the four actions Tom and the NFL has presented as luck events thus I find it difficult to deny luck in the game of football so how should teams handle this how should people handle this as I was making this video I struggled to find many coaches talking about luck but it wasn't difficult to find something else luck is when preparation opportunity this phrase has been beaten into the school of anyone who's played you sports but in a way it is admitting to luck you obviously can't control everything and that's good advice everybody would agree with that unfortunately there are still many that deny the existence of luck and I think that that only increases the likelihood of being fooled by luck being fooled by by Randomness for example how many players and coaches have received massive contracts largely due to luck after quarterbacking the Giants to a playoff win Daniel Jones went on to receive one of the biggest contracts in history that playoff win came against a team that had been quite lucky a team we previously spoke about the Vikings the next season didn't work out very well for the Giants would Daniel Jones have received that contract at the conclusion of this season not how many GMS hit the lottery on a single draft pick and look like no stradamus how many GMS get fired because they have a long-term strategy that the owner maybe just didn't have patience for I did a video on this by the way the NFL draft is wildly inefficient and I recommend you check it out right experts players and coaches don't deny the existence of luck they identify it and embrace it they don't overconfidently create a narrative off of a single game they create a measurable strategy for the series of games because if you can measure something then you can look to manage it hope you enjoyed this video make sure to subscribe if you haven't and take care

Channel: Michael MacKelvie

Views: 191,748

Keywords: NFL, NFL Analytics, Sports Analytics, Football, kansas city chiefs

Id: Zfin10hS0YM

Length: 16min 48sec (1008 seconds)

Published: Thu Apr 04 2024