Perhaps the most surprising part of my results is that the quarterback MWAR is relatively low, only influencing -0.14 games in comparison to -4.91 that was concluded by PWAR. There are many intangible traits that influence success in the quarterback position such as confidence and chemistry, which are hard to replicate in a video game simulation. Countless NFL quarterbacks who display superior speed, arm strength, or other physical attributes have failed to shine in a game situation, whereas players such as Tom Brady, arguably one of the least athletic quarterbacks of the 21st century, have thrived for the better part of 15 years.
More so than any other position, quarterbacks must bring their mental ‘A’ game if they want to be successful. When a quarterback wants to make a pass, he must read the defense properly, ensure that his receiving player is running in the correct pattern, and calculate his throw so that his receiving player is able to catch the ball without any opposing player touching it. If any of these mental steps are performed incorrectly, it is likely that a bad pass or interception for the other team will occur. Although Madden attempts to replicate this aspect to the best of their abilities, it is implausible to think that this skill set can be truly measured by a video game. I believe that the game gives the benefit of the doubt to an unrealistic number of quarterbacks when determining their ability to perform mentally, though I do not blame them as the purpose of the video game is meant to be an enjoyable experience for gamers, and not to be a perfect representation of the NFL.
The runningback position, on the other hand, requires more physical prowess and individual talent to be successful: in real life as well as in Madden. In comparison to quarterbacks, there are far fewer variables in play when a runningback is given the ball. Speed and power are the two factors that influence success for runningbacks, and can be numerically quantified much more easily than mental strength. The results show than the replacement runningback ran for half as many yards as the starter during the regular season, largely because the replacement player was not nearly as fast nor as strong as the starting player. Due to this reason, MWAR values for runningbacks were quite significant, as it is difficult to replace raw speed and power.
For defensive positions, the left tackle is has shown to be the most influential position for generating wins in MWAR as well as in PWAR. Traditionally, the left tackle position has been speculated to be the most important position as they protect the “blind side” of quarterbacks when they are looking downfield. This position has been difficult to quantify the value of, however, because there are very few statistics accredited to this position. In Madden, the only statistic recorded for left tackles is ‘Sacks Allowed’, which is the number of times a player allows their opponent to get behind them and tackle the quarterback before they can make a pass. It makes sense that, if a quarterback is sacked fewer times throughout the game, that their team will be more likely to generate more offense and ultimately wins; Madden simulations only confirmed this notion.
In a similar fashion to left tackles, the value of linebackers and cornerbacks has been difficult to determine numerically. For linebackers, PWAR concluded that they are worth approximately -1.17 game whereas MWAR found that they are worth -0.12 games. Although not many conclusions can be made from MWAR Linebackers due to a high pValue, I think that the true value of the linebacker falls somewhere between the calculations of PWAR and MWAR; it is not quite plausible that linebackers have almost zero impact on wins, nor is it believable that they are worth more than one game when every team rotates multiple linebackers throughout the course of a game.
Conversely, MWAR estimated that cornerbacks are worth -1.22 games in comparison to -0.091 games from PWAR. Cornerbacks are trusted to guard the opposing teams receivers, and stop them from catching passes. Above all, cornerback are expected to prevent “big plays” from opposing teams, thwarting attempts made by the quarterback to complete a very long pass to a receiver. In theory, if a cornerback can stop the opposing team from gaining yards in big chunks, it can considerably increase a team’s chance of winning football games. Though losing an extra 1.22 games may be generous assumption for the loss of a starting cornerback, that number seems to be more realistic than the near zero value computed by PWAR.
As a whole, this was a fun and enjoyable project, and I have not eliminated the idea of revisiting this idea to conduct more simulations and expand on the conclusions that I have already generated. Thank you for visiting this website!
Previous: Individual Player Analysis