Dowdle is a High Variance Rusher

What do you think the flaw in the argument is again? Because it's unclear to me what your objection to succ% is...beyond handwavy "empty rhetoric" of course.

Please be specific. And if you can, please type it slow so us numbers folk can understand.

Maybe start with what you think my argument is? Then you can point out why you think it's flawed? Again, please be specific on why your objection "points out the flaw" in my argument. Shouldn't be hard for a modern day Cicero, right?

Because last time you said you pointed out the flaw in my argument, you said you didn't know what my argument was when I asked you.

I also think it's funny that the other day you were criticizing someone else for using appeal to their own authority as a negotiator. Now you're trying to win this argument by telling me how great you are at rhetoric. I'm an experimentalist, so I'd prefer that you show me instead of telling me.

Also, have you ever thought about how subjectivity total yards is? My guess is that it's not easy to find 5 stats in the NFL that are completely objective...if we're super anal, there may not be any.

My first thought for a completely objective stat was pass attempts. Then I remembered this play.



Which is why the NFL rule book is enormous and I assume crafted by highly trained lawyers. Because everything in the NFL is subjective.

Aside: This is why I dislike the "football move" rule for receptions and much preferred the "you have to keep possession when you go to ground" rule. Even though it meant that James didn't catch that ball...but if we ran that same play now, it would be a catch.

The keep possession rule was more objective than the football move rule, despite still having some subjectivity. It's much easier to understand. And it's also how you'd play if you were just throwing it around with your friends.

 

I missed that the first and only time I watched it. Gonna have to go back and look again. I've seen several different versions on various streaming platforms....got one that stands out? Extended? Directors? or whatever.

 

Is this all just a really long discussion for a certain someone to get back round to how Harris was/is a better RB than Warren?

 

Honestly I watched whatever version was on HBO, Showtime, or whatever movie channel I 1st watched it on, but because of being such a fan of John Carpenters the Thing, which I own about 3 different copies of, and have also on my DVR, I noticed it , and pointed out to a friend of mine while watching Hateful 8 the 1st time. The soundtrack/score was the 1st clue.

 

Modern day Cicero had me rolling with a great belly laugh.

 

An interesting comparison from the Depot:
https://steelersdepot.com/2026/07/study-a-key-difference-of-steelers-new-running-back-duo/

Shows that Warren ranked 4th in forced missed tackles on running plays by RBs (despite having the fewest number of carries in the sample).

Dowdle had 25 more carries than Warren and forced 41 tackles.

On passes: Warren had one more reception than Dowdle. Warren forced 24 missed tackles. Dowdle forced 7.

I think both will be good (but not great) backs for us. I hope that Warren gets the larger workload of the two because I think he's the better back. But I won't be too upset as long as it's around even (provided that they keep giving similar results as they have in the past).

 

I’m anticipating George Pickens being added as well

Some people think if they say the same thing 37 times while ignoring the facts it helps them gain some lost credibility

 

I just explained this. I have explained the flaws in success percentage before. In your rush to try to falsely portray something subjective as objective, you contradicted yourself. You claimed I didn't understand the word, then admitted that your pet statistic is subjective. That a bunch of statisticians decided they have refined the statistic enough to like it doesn't change that it is subjective, and therefore, inferior to objective measures.

Now you are trying to argue that poor interpretations of the rules on a specific play are the same thing as using the statistics that result from one game or many games. It isn't.

Still waiting for you to admit the claim that I dodged anything was pure nonsesne.

 

Almost every spot of the ball is subjective. I think the exceptions are: touch backs, XPs, adjustments for penalties (though the "previous spot" would have been subjectively determined in these cases).

There is no objective way of determining where the ball is at the end of a play in many cases. And it would make the product aweful if we reviewed the spot on every play...which would be closer to an objective measure in cases where we can see both the ball and whether a player is down / in bound in the same frame...probably multiple frames because you also have to figure out where the ball is relative to the yard lines. It's actually pretty frequent that the ref on the field can't even objectively determine when a play ends. And that doesn't even account for how subjective it is when they call / don't call holding (which happens to some extent on basically every play). Pretty much every decision made by the refs in real time is subjective. Also true in the replay booth, but they have more information and time so they're probably more accurate in large samples.

This means that every yardage stat is subjective.

But just because these stats are subjective doesn't mean they aren't useful (kind of like the George Box quote...I think made famous by Feynman). But just because stats aren't the exact same thing as "ground truth", doesn't mean they aren't "useful".

Which is the point that you keep missing. Others above have suggested you may be intentionally missing this point to avoid admitting that you were wrong in a previous argument.

I don't think it's true that you have "explained the flaws in success percentage before". I believe that you've only said that it's subjective (and implied that means that it's arbitrary or at least not useful). I think you've explained why you think it's subjective. But never why it's not useful. If I missed it, I apologize. I'd be happy to be proven wrong if you have a link or something. Or you could just write down what you think the flaws are and we could try to discuss them.

So I think re: succ% you're arguing against a straw man. You seem to think that I think it's (1) perfect or even (2) good by itself. My argument for the benefit of succ% isn't that it's either of these things.

As I've said to you many times, my argument is that: I think that combining YPC (how many big runs a player has) and succ% (how consistent the runner is) gives you a much better sense of a back's performance (in large samples) than any single metric alone.

I've believe that you've argued that YPC is a good metric for consistency. Again, you got an object lesson (it's funny, I always want to say "abject lesson") in this against the Lions last year. And Budda's post above shows that YPC is a strong function of the yardage a back gets on the top 10% of their carries. It doesn't tell you much about the other 90% of their carries...because those are basically the same for all backs (which Budda also showed us in a sample in this post while comparing the distributions for Dowdle and Warren).

Again, when we first starting talking about this you made the argument that it was a fringe stat that wasn't available on reputable sites like PFR. And I would have agreed with that 5 years ago. But you need to "update your priors" here. Because these arguments are not longer valid.

I believe you've also argued that Najee 2 years ago was better than Warren last year because Najee got over your subjective benchmark of 1000 yards (1043 yards on 263 carries). But Warren only got to 958 yards (85 fewer yards and 52 fewer carries).

Re: Dodging. I would prefer to remain honest in my assessment of your previous post. You may disagree, but it's my honest assessment.

Also, FWIW while I disagree with you on this I'd always be happy to have a drink with you (or anyone here) and talk Steelers. It's no fun if we agree on everything

 

What is a variance rusher? Lol wow!

 

Man, people think I get wordy. I would still like to see you acknowledge that it wasn't a dodge to focus on pointing out that you contradicted yourself when you mocked me for calling success percentage subjective, then admitted that it is subjective. That isn't a dodge. I don't dodge. I do occasionally get bored or annoyed enough to stop reading convoluted nonsense, though.

Sure, subjective statistics can be useful in some situations, but objective ones are better. Despite your misleading attempt to muddy the waters, that is exactly what yards per carry and total yards are: objective statistics.

Your false claim that all statistics are subjective is utter nonsense. You keep talking about how the statistics are recorded. That is not the same thing as what we do with those numbers once they have been recorded. You are comparing apples to oranges. Yards per carry and success percentage are both oranges in this metaphor, both measures applied after the statistics are recorded.

I don't ever remember dismissing success percentage as a fringe statistic. Perhaps there was that implication in a post at some point. I did explain how these people define success is at the heart of its subjectivity and potential inaccuracy as a measure. The way I have seen it explained, it allows for a bit too much focus on what the whole team is doing, not just what the back does. The team affects all statistics, but some more than others, and the effect is too high with success percentage.

 

I asked that on page one

 

sorry I missed it, I just find it funny what people will post.

 

A low variance player will give similar performances game to game or touch to touch. The OP was looking at game to game stats (something like succ% gives you some idea of variability vs. the goal of the play on a play-to-play basis).

Looking at an average performance of this player will give you a reasonable prediction of what they'll do next game. I don't have the data in front of me, but Bell might have been a low variance RB (when healthy). I don't remember him getting a ton of big plays, so his YPC probably wasn't driven up by big plays as much as a guy like FWP would have been.

A high variance player will have some amazing games and some amazing games and some really terrible games. Looking at the average performance of this player won't help you much if you're trying to figure out what they'll do in the next game. For Steelers, I think of Ben as an example of a high variance player. Where you never knew if you were going to get "good Ben" or "bad Ben". For RBs, FWP might have been this guy...but carry to carry. He did get a lot of usage, so his game to game performances might have been similar?

 

I just didn’t want you waiting for an answer

 

In two games last season, the loss to the Bengals and the victory over the Lions, Warren rushed a combined 30 times for 270 yards, nine yards per attempt. In his other 14 games, he carried the ball 181 times for 688 yards, only 3.8 per carry.

By the way, if you lump both Bengals games and the Lions game together, he had 40 carries for 332 yards, an average of 8.3 per rush. In the other games, he had 171 carries for 626 yards, only 3.66 per carry.

Sounds like high variance to me. To be fair, Dowdle did almost all of his damage in a five-game stretch from Oct. 2 to Nov. 2 and had a lousy season outside of those performances, but let's not act like it is only him.