You cannot compare expected points on runs to expected points on passes. They are two completely different things.
It obviously isn't per play on the y axis, otherwise the x axis would not be there.
Also - I reiterate. The statistical model used to generate this graph has Beasley as adding 40 points. He scored 3 TDs all year. Yet, you are telling me that this model is accurate?
Statistical models are PREDICTIONS. To assess the relative value of a model we compare the PREDICTED outcomes to the ACTUAL outcomes.
The actual outcomes do not match the predictions ... in fact they are way off. The only conclusion in that situation is that your model does not accurately represent reality.
But please - keep pushing your agenda - now supported by meaningless statistics.
Man, this analytics revolution has people thinking they know something when they really have no clue.
I'll reiterate: You don't gain EPA just from TDs, just like an 8 yard catch on 3rd and 7 adds more EPA than a 15 yard run on 3rd and 20. I already stated this, and I guess it didnt click with you.
You saying TWICE now that since Beasley had 18 points scored, his EPA can't be 40 highlights you do not understand the graph.
You can compare runs and passes, it's in the graph.
This graph isn't saying Beasley is a better player than Zeke. This graph is saying last year Beasley far and away helped to team win more than Zeke on their touches. This is partly to do with play calling, partly to do with the fact that running backs don't really help you win, and partly because of performance. That's all.
The only conclusion we can make is that you can't correctly digest the graph. Unless of course you think you know way better than Vegas, which you obviously do not.
This statistical model IS NOT A PREDICTION, IT WAS AN OUTPUT AS RESULTS. It's not predicting what will happen in 2018, IT ALREADY HAPPENED
I will end the conversation here.