He has absolutely no idea and his player models are often wildly inaccurate yet everybody posts them here as if they are gospel.
Can you provide some examples? I haven’t seen every card he’s ever had so idk. There’s been players with outlier seasons that have been anticipated/guessed (more easily with older players than young developing guys) a potential drop or breakout.
Jack Hughes, Nichushkin. Seider is the kind of guy you look at his underlying numbers as a rookie and you feel confident he can become a top Dman despite the raw numbers of how many goals he was on for, or how few points he got on a bad defensive team with less scoring depth.
His model isn't even his model. The source of the model knows what he's talking about though, and the source of that model also knows it's got little predictive value but you'll never hear that part mentioned in a Jfresh tweet.
That’s literally true for all stats including points, and what he does with those stats is the same thing everyone has been doing with stats forever lol. They’re all measure of something that happened in the past. We want to talk about the future because it’s fun but nobody can see the future, so we try to use recent past history.
Someone can score 80 points this year but that doesn’t mean they’ll automatically 80 next year or that their production can’t grow or drop. Maybe the created a ton themselves as a set up guy or maybe they’re a finisher who needs the former. Maybe they boost their offense indirectly with stellar play away from the puck/defense that their teams transition back to offense more often seemless, but how do we know unless you watch literally every game like almost nobody? Some players stay level, some improve, some young players improve one year then drop back down the next etc.
The point of those other measures is there’s so much more to hockey. It’s a game of many many small battles and races, small decisions like positioning or do I attack or sit back etc. The micro stats are where you really see where the overall numbers are coming from. Two players might have similar overall 5v5 defensive impact but for different things (entry defense, controlled exits; maybe he suppresses shots because he’s insane holding the ozone blue line etc). It can still come down to role and how their coach uses them, but that can be said with points too again (how many minutes? With who? Was he on the powerplay? Does their powerplay coach suck so the whole team does or does this individual player drag it down when he’s on vs off, etc)
Everybody wants to talk about next season and the future. That part of the fun of sports. Strong underlying numbers can often be a good sign that a player is improving away from the puck and winning lots of the small battles (Mark Stone, takeaways), or even super effectively just using positioning to steer opponents puck movement (Bergeron, Pulju). *
but like I said, it’s still trying to guess the future using the past. Pulju scored a lot more before getting covid and injured. Do we know he’s going to score 40/less than Yamo again? Jfresh says/guesses Pulju is better driver (ie does the little stuff well more consistently) because the numbers say he was last year. They’re both still young however, and maybe Yamo majorly improves as young guys do and or maybe Pulju comes back next year demoralized or having some issue and doesn’t replicate his defense/two way play.
Analytics is especially great for closing the gap in offense vs defense measurement. Defense used to be so much reputation based, not to mention a great team defensive scheme with full commitment can mask personnel deficiencies, so breakdowns to isolate an individuals impact is useful. eg the Oilers did better ratio of chances or shots generated and allowed w Pulju on the ice vs when he was swapped off a line for a different winger, compared to Yamo, and if I remember correctly he seemed to majorly boost zone exit efficiency which isn’t nothing. Means little in the conversation of Pulju vs Yamo for next year tho. They both paced basically the same points last year but development isn’t linear or consistent.
I mean, that's all well and good but wtf is that statement? "occasionally produces next to superstars but hard to say if he is driving the bus".... What players exactly would we expect to be driving the bus on a line with McDavid or Draisaitl? Also players that drive the bus don't make 3.1M.
Nice conclusion Jfresh
I think the wording is just weird. There doesn’t have to be one “driver” and therefore the other two guys arent automatically passengers.
If you don’t want to read the above wall of text, the main part here is that hockey is a series of many small battles, races, and quick decisions on things like positioning and aggressive/passive choices. Analytics measure play away from the puck as well as with it, and try to determine the connection.
A “play driver” is generally a player who does many of those little things well enough consistently/efficiently enough that his team has an easier time getting the puck back and or getting it up the ice, and keeping the puck moving around the O zone or at least disrupting a teams breakout attempt whether with his stick (Mark Stone) or with his positioning (Bergerons IQ being the prime example.
A player who does all those things well to some extent; or just a few of those but at an extreme level would be an elite play driver. A player who does a few pretty well or one exceptionally well is a good driver.
A line with three play drivers (or 4 or 5 w Dmen) is more likely to spend more the game in the other teams zone than a line with only one driver. Basically sorta possession, which is a defense stat primarily and a “scoring potential” secondary in the sense that you naturally have higher odds of scoring the more often you have the puck, but you would still need some finishers. Without, you’d still have great defense impact. This is the Avs top 5 man unit in a nutshell.
Using Pulju as the example again, his centers saw better ratios when he joined their line, far as shots and how badly they outscored opponents. Even tho he wasn’t scoring a lot himself, his linemates were still scoring but they also weren’t allowing as much. To whatever extent some of a players lack of finish can theoretically be mitigated if he well enough makes sure his linemates get the puck back up the ice much more consistently, that they themselves can create more chances.