This is a prime example of how you use "statistics" to dissemble. Yes - Dostal has had bad games.
So you're going to focus on those 3 games (not back to back) rather than the overall numbers (which are very good). Cherry picking at its finest, particularly when you consider that a goalie can play well and still give up a lot of goals (particularly on recent ducks teams).
And then you're going to take a 10-15 game sample and interpret it w/o any context - the exact context I provided which goes against your narrative (i.e., the atypical structure of the ducks schedule, with fewer games of the stated period, made playing Dostal a lot the correct choice). Notably, you didn't address this.
By analogy, I could claim that playing Dostal as the ducks did is "obviously" a good thing because his save percentage is .922 which I believe is top 5 in the league. He's also leading the league in goals saved metrics. I mean, the results speak for themselves, don't they? The ducks are obviously doing the right thing - after all its recorded history that Dostal has played VERY well this year before Gibson returned!
You think your "numbers" are dispositive. They are not, particularly because you cherry pick the numbers to reflect your narrative. You are selecting the "recorded history" which fits your narrative. This is not the first time.
Interpretation of raw data (as you attempt to do) has nuance and doesn't dictate solely one clearly correct conclusion. That is why "statistics lie" and your charts convince no one - most people understand what your doing here.
Unfortunately, you're doing a bit of the same thing you're calling out here.
From the start of the season to the day before Gibson's return, Dostal played 11 games. Of the 72 goalies who played
during that span, 5 played in 12 games and 7 played in 11. So, a little more than a third of the league, all good so far. No other goalie who played that number of games had a backup who played fewer than 3, while Reimer played in 2. All good, lines up with your point about the Ducks playing fewer games.
But your assumption that the lower number of games spread them out more isn't accurate. The Ducks had fewer games because they started latest, not because the schedule was less dense. If you
condense the timeframe to only include the time after the Ducks started play, Dostal's 11 games are matched or exceeded by only two goalies: Saros and Vasilevskiy. It's the
other teams' goalies who were benefiting from a more spread out pace. Also worth noting, during that stretch, he faced the fourth highest shots/60 and sixth highest high danger shots/60 in the league. You have to go down to 9 GP to find anyone whose workload was even close.
Dostal does need to learn to be a starter, but this is not a wholly reasonable learning experience. This over a longer timeframe is exactly how Gibson got Gibson'd. And the fact is, Dostal absolutely
did crater at the end of a stretch where he played more games with a higher workload than nearly any other goalie in the league. That's not cherry picking, it's a valid data point. If it were three random games being held up as representative, that would be cherry picking. Pointing out that Dostal crashed at the end of a high workload and bounced back after Gibson relieved some of that workload is perfectly valid analysis of the underlying stats.
The sample size is too small to
conclusively prove anything, for sure. But the argument is more reasonable than you're giving it credit for.
