Man if I hadn't suspected Drance was low-key shilling by this point, I think the above would've been a huge red flag. I saw his tweet and was curious if this was true so I decided to dig deeper on naturalstattrick. I know this is the media thread, but I'm gonna bring up a few numbers just to illustrate my point:
2018-19 through 2019-20:
| 5v5 TOI | CF% | FF% | SF% | xGF% | SCF% | HDCF% |
Schaller-Beagle-Motte | 138:36 | 48.66 | 53.33 | 54.86 | 58.55 | 50.86 | 66.00 |
Beagle | 853:46 | 40.80 | 43.21 | 43.32 | 43.89 | 41.82 | 45.75 |
Schaller | 735:57 | 44.52 | 45.30 | 45.24 | 46.72 | 43.75 | 47.94 |
Motte | 900:22 | 44.04 | 44.73 | 44.56 | 43.37 | 43.01 | 44.38 |
[TBODY]
[/TBODY]
(all numbers from naturalstattrick.com)
While it's true that their numbers look oddly good as a trio, that's also from a grand total of... something like 13 games combined over the last 2 seasons.
So Drance isn't
technically wrong here, but as a known stats guy you'd think he would at least admit to the one glaring thing about the numbers that stands out almost right away: potential sample size issues! I mean, we have 3 players who've all put up mediocre to horrible possession numbers across the board, over hundreds of minutes, who've suddenly
vastly improved when playing together over a mere handful of games? Yeah I'm not skeptical here at all...
This whole thing kinda falls under a dubious sort of "grey area" reasoning and just seems a little too disingenuous to me to be comfortable. It definitely feels a little off knowing the kind of content he used to put out in the past... so I'm not gonna lie - I'm a little disappointed.
What almost makes it even worse is how the vast majority of people in that Twitter thread just lapped it up and took it as gospel outside of a few woke ones. I'm surprised no one thought to ask what kind of data he was drawing his conclusions from.