(And here is the direct link to the chart. May be able to view it better.)

This is basically a collection of small sample sizes, but taken together it makes a pretty good sample size, especially for someone like Ryan Strome, who didn't have regular linemates throughout the season. (Some forwards had around 700 minutes with another forward, but the other forward had 200 minutes without him, so the sample size effectively shrinks to 200, despite all their time together-- such as Kucherov/Palat/TJohnson, because they hardly spent time apart.)

Reasons Strome's chart is very compelling:

1) Each of the six forwards Strome spent the most time with, the Isles scored better with Strome on the ice than off the ice.
2) The two who appear close in WOWY production (Lee and Tavares) had drastically better zone-starts without Strome than with him (Lee 64% to 52%, Tavares 71% to 51%).
3) Even each of the 5 who spent little time with Strome did better with than without, according to this stat. (Bailey+Strome blue bubble is "off-the-charts" in their small sample size.)
4) Overall Strome had 49% zone-starts at 5v5 whereas these forwards (weighted by TOI together with Strome) had 53% without Strome.
5) Strome was also respectable for the defensive stats: GA/60, SOGs against/60, CA/60, and FA/60 were all near average, among NYI forwards.... Strome was no Bergeron, but it wasn't like he ignored defensive responsibility to contrive more goals.

Other interesting things about this chart:

Wrap your head around this: At 5v5 with any combination of Strome + another forward on the ice, NYI scored goals at a better pace than they did with Tavares on the ice without Strome (in general).... Though the last 5 combinations on the chart were very small sample size and somewhat fluky that each were above that....

Part of this I'm sure is due to opponents focussing on Tavares on 5v5, just as Stamkos's 5v5 production was limited while the TBL 2nd line tore up the NHL, but for Strome to be in the same conversation as Kucherov/Palat/Johnson is great, after watching the playoffs. (Top seven in on-ice goals-for per-60 at 5v5 were Kucherov, Palat, TJohnson, Strome, Benn, Nash, Tarasenko, in that order.)

Notice red bubbles of Lee-Tavares-Bailey.... That was pretty much the only NYI elite line without Strome, in terms of goal-production at 5v5.... (Okposo hardly had any TOI with Strome, so he didn't make the chart, but his goals/60 without Strome is 2.59.)

Next best forward for on-ice goals/60 without Strome: Grabner. And he did so with 38% zone-starts without Strome.... Grabner and Nelson combined for 2.46 GF/60, 62%GF, and 55% CF together, despite 35% zone-starts. Grabner+Kulemin similar: 3.11, 75%, 53%, with 37%.... Perhaps the most successful "shutdown" line for Isles this past season, in a small sample size. (Better than the best 4th line in the NHL, I'd say-- and I do respect what 4th line did for NYI this season.)

Grabo+Strome were through-the-roof good: 4.69 GF/60, 67% GF, 78 CF/60, 58% CF, with 46% zone-starts. Only 128 minutes together, but 58%CF with 46% zone-starts is spectacular.... Grabo also did great with Tavares, Okposo, and Lee in small samples. He's an intriguing option for 2C or LW1/LW2, depending on whether Nelson is LW or center.... In other words, Isles don't have to play Strome at 2C, especially if he can perform similar to this again next season at RW at 5v5.

(More charts coming, to compare to this one of Strome.)