This is solidly under the “dataset limitations” issue. We don’t have any of the 2-2 or worse GWPs, which is why I make so many disclaimers throughout the article about the numbers we are using. So we don’t know the conversion rates from the decks that start a daily to the decks that finish it at 4-0/3-1, and we don’t know the (obviously lower) GWPs of the decks that go 2-2 or worse. One day we can hopefully expand the analysis to include those stats, but we need way more data that Wizards currently doesn’t make available.

Thanks! Let me know if there’s anything you think the community would like to read more about!

About the deck, I have it in paper and play it sometimes, but am not nearly as good with it as one should, and this is one of the reasons I think are responsible for the deck to not be played more: it’s very hard to pilot. I would compare it to Bloom Titan, and say it’s somewhat inconsistent. When the draws help, the deck is almost unbeatable, but the weak draws are very problematic. However, it is the Modern deck with more inevitability, it simply wrecks Midrange and Control. Combo, unless blazing fast, fall apart as well. The matchups against Twin and Junk are very good, and this is something I believe has correlation with the deck performance, as those two decks are among the most played.

The weakness of the deck is well-known, aggressive strategies, but it is surprinsingly good against Red based decks. Wurmcoil Engine and Platinum Angel are really hard to deal with. To say the truth, the weakest matchup are those that offer a fast clock with access to Path to Exile, usually Wilted Junk and Zoo.

All that said, I think the only thing keeping U Tron from reaching the top stages is a great finish on a major tourney like a GP.

Glad it was helpful!

One thing I didn’t do was hypothesis testing for the different mean GWPs. This current analysis constructed a confidence interval for the overall population (a pretty broad one), but didn’t then construct a similar interval for each deck. Although MUT has a high GWP, it also might have high variance, and therefore its confidence interval might overlap with the population’s. If so, there wouldn’t be a significant difference between Tron’s GWP and the overall one, even though the average GWP was higher.

Definitely a followup article idea!

This is actually a really helpful explanation of the data and I’m looking forward to content like this in the future. Maybe I should switch over to mono-U from RG in paper magic just to try it out…

I’m not 100% sure it’s “ignored”, and that’s one of my more speculative points here. But it definitely is underplayed, at least if the 4-0/3-1 finishes are any indication (that gets back at the dataset limitations I mentioned earlier). The only reason I am inclined to believe it is ignored is because some Mono U Tron players have said as much in the past. In that sense, it’s a case of the quantitative data showing evidence of a qualitative phenomenon.

Hopefully we see more Mono U Tron in the future though! It’s a great deck.