Soooo, it’s been a month and a half, and I’ve had some time to think about the above, but more importantly (in terms of addressing it), to try more classes/runs and get a better idea of the variance in the mode. This is still anecdotal, but as someone who played a bunch of Dungeon Run/Monster Hunt, I think I can do a good job of comparing them to the current Dalaran Heist.
The most fun version of this, by far, was Monster Hunt. This was for 2 reasons:
- The classes were very unique in terms of the play styles
- The variance was (relatively) low.
I could reliably beat any character’s run more than half the time. Toki was over 75% if I remember correctly, with more than ten runs. Most characters had more than 1 optimal (or at least very good) treasure/play style. All of this just made the mode a lot of fun.
Dalaran Heist has some extra uniqueness, due to the anomolies and alternate hero powers, so this is nice, but any additional fun/uniqueness from this is severely limited by the next point.
The variance is extremely high. And this leads to much larger numbers of games where player agency is simply not a thing (i.e., there just wasn’t any available route to a win for that run/game). OR, alternatively, where you would have to play extremely poorly to have a chance at losing. Neither of those are very “fun”, I imagine, for most people.
If you start off with Robes, and draft around it, you should be extremely favored to win in most circumstances, and your choices will usually be at least somewhat obvious. This is often boring.
Now, it is still possible to lose when playing perfectly, if your opponent gets a stupidly good draw, but those games are not fun either. There is very little “in-between” area where the AI gets a decent draw and your in-game choices actually matter, as long as you understand the basic game mechanics.
If you don’t start off with Robes, then the above “auto-win” factor is gone, but, the probability of auto-losing to a good draw goes WAY up. A win is still very possible, and I won’t claim otherwise, but it feels (to me) like significantly less than a 50/50 chance per run, which just gets tedious since they are 12 games long. Player agency is still a thing here, since your chances are probably 2% with “halfway decent” play, and, I dunno, maybe 40% with “really good” play, but that doesn’t seem rewarding enough to me in “non-Robes” runs.
All of this makes the replay value for Dalaran Heist much lower than, say. Monster Hunt, which I played a LOT.
Dalaran Heist did split up into “normal” and “hard” modes, which is a huge point in its favor. But “normal” mode was rudimentary, which is OK once just for flavor, but without any real challenge I won’t be doing it a second time. And “hard” mode falls victim to the extreme variance mentioned above.
I liked a lot of the additions they made to try to keep runs fresh, and the deck-mod option was universally liked, and also somewhat skill-testing. But the variance in treasures, and possible stupidly-good draws by the AI, mean that I really want a better experience for any future solo modes.
They could have made “hard” mode so much better by play-testing a few more times, acknowledging the variance, and then toning down some of the bosses and the overpowered treasures. This would have allowed for more play styles and decks to be actually viable, instead of plain terrible. (For example, just ONCE I would like aggro to be a viable option in these things, but the bosses practically always require a value-oriented approach. Same for token decks.)