I did, but we seem to be going around in circles. Last post here, I swear.
This chart below shows individual zone starts. It doesn't say that Sven is getting OZone starts 44.8% of the game. It says that when Sven is deployed, it's in the OZone 44.8% of the time, and NZone or DZone the other 55.2% of deployments. It does not quantify at all how often he's deployed overall so it's hard to see the relevance to the argument at hand.
The PITB article does quantify.
The data doesn't lie, as was pointed out a few times. It just needs to be understood and telling me to get lost doesn't help me understand.
But you don't understand it, you just key on one idea and turn it into a huge pile of misinformation.
That idea being that the data is faulty because it doesn't tell us how often somebody is deployed overall.
First of all, the total % of deployment is unimportant because you can look at that in terms of TOI. What you are completely ignoring is that you can choose to look at it in a broad or specific context. Neither context supports the spurious claims you are trying to make.
For example in your quoted chart, it shows that the TOI for Dorsett is 99:56. Jannik Hansen's in the same # of games is 96:34. Now that we've chose to look at TOTAL ICE TIME, we don't need to think about the total % of the game that Dorsett or anybody is deployed.
That number (zone as % of deployments) is meaningless anyways because if you further broke ANY player's total deployment in the context of 60 minutes it's going to look miniscule. For example Edler plays 20/60 minutes which in terms of the whole game is %33. If we look at what % that is of the blueline, well there are 60 minutes and two guys on the ice at all times so that is 120 minutes for all blueliners. Edler's 20/120 is ~16%. If we broke that down his zone starts that 16% is going to be fragmented into 3. At this point the numbers aren't useful, simply because even a player who plays a lot doesn't get a huge amount of TOI% or shift% relative to the entire game. You could make it better by taking ratios of the each blueliner but now you are having to manipulate numbers that are already fairly well explained by TOI. Let's say Hutton only played 15 minutes, which is less than 20 minutes. There is no need to break this down into whole game % since 15 vs 20 is pretty self-explanatory. And it is automatically selective because you're only comparing defencemen.
So your first premise is faulty.
Let's get back to something a lot simpler. In the context of total TOI in what you've quoted, again Dorsett is 99:56 and Hansen is 96:34. So right away you see an issue. Dorsett plays more than Hansen. This is frankly ludicrous.
Now let's look at o-zone starts. We've already seen that Dorsett plays more than Hansen, which is stupid. Now you see that on top of playing more, Dorsett is deployed in the ozone 38.6% of the time and Hansen is 37%. Again there is no reason for such a disparity, and nobody cares about what these ozone deployments are in terms of total deployments. Because they have comparable ice time which means that ozone% is comparable. We have simple stats telling us a story that you want to manipulate in order to tell us the same story with smaller numbers (because if we did bother to figure out total deployments it will come out the same because these are freaking ratios).
Please stop, you are not making yourself look good here. This isn't even a 'fancy stats are limited' argument, these are very simple statistics that tell a very straightforward story. If the 4th line was played less in 1-goal-down situations (say around 80 minutes), nobody would care if Dorsett was started in the o-zone often because then you could make a 'sheltering' or 'hitting' argument.