Part 12: Interlude 1 (Evaluating Strategies)
Interlude 1So I've completed Chapter 11 several times today using different strategies and attempted several others that I stopped partway through when I realized they weren't going to work. I believe I have discovered the best one feasible and I'll post that later on today. For now I want to draw attention to someone else's strategy:
https://www.youtube.com/watch?v=-YOjcmO7jBw
Ultimately, I think my own approach is superior in that it depends less on good luck or a tough Matthew, but it was this guy's video that got me to completely rethink what I'd done in previous runs and come up with my new strategy.
And that's a big part of what I'm trying to do here: give anyone else interested in doing a max ranking run the tools and advice they need to formulate their own great tactics on each level. That's why even though I said I wanted this LP to be part guide, I don't intend to post what I did move by move: instead I want to show an outline of what I did and tell you what strategic considerations made me think that was a good idea.
Besides crediting this guy for a strategy that inspired me, I'd also like to try to explain to you what I think HE was thinking. If you want to be good at this game, it's not enough to read or watch what other people did. You have to think about why they did that and whether you can see a better way to handle things.
So enough rambling, here goes my analysis of why he did what he did:
0:10 - The big choice is top route or bottom route. Top route is faster, bottom route is safer and more XP. He goes top because his goal is maximum speed. And to do that efficiently, you need to break the wall on turn 1, so Hector must use the hand-axe from range.
0:23 - That soldier needs to die and only Hector can do it. He wants to conserve his precious Wolf Beil. He could pull Matthew back and have Hector attack with the handaxe with impunity, but his main concern is speed so he charges in.
0:30 - There's the big twist! That's a hard move to explain because you'd think he'd either have Matthew attack the archer or open the door. What he's doing is manipulating the AI. Move Matthew into literally any other space in that room, and the archer will move to block either Hector or Matthew. But if you put Matthew in the top middle, the archer gets nicely out of the way. Again, he's thinking only of speed rather than survival chance or anything.
0:40 - Hector needs to be gunning for Wire at this point, so he moves full speed. Conveniently, this will also block the archer so the archer will be forced to eat a handaxe to the face if Matthew moves past Hector.
0:43 -Matthew is safely out of the way and can unlock the door for Hector next turn. Critically, he has enough HP to take one more hit from the archer if need be.
0:55 - He leaves Matthew in harms's way and charges Hector again. It is crucial that Hector did not stay behind and attack the archer. The instant the door is opened, the bottom enemies come running. If Hector does not run through the door immediately they will either block him in and waste his time, or gang up on him as he fights the boss and kill him. So Matthew must be abandoned.
1:10 - The archer must die or Matthew will be killed. More importantly, Matthew MUST attack from the top of the archer, not the left. This way he will be close enough to the currently locked door that he can run into its space next turn. This guy understands the thief AI and knows that the thief will stupidly open that door instead of making its getaway with the jewel. This gives him the chance he needs to steal it.
1:20 - I've thought about this a lot and it looks like a huge blunder. Putting Hector 1 south, 1 east of that position would be far superior. Why? One fewer archer can shoot him. Furthermore, the soldier cannot block him. If Hector missed that soldier, he would have been prevented from reaching Wire. You can see that even though he lucked out, he's horribly injured. A single hit from Wire will kill him and Wire gets 2 shots with 60% odds. There is one possible advantage to this space though: if he hit both archers (about 36% chance) he would immediately level up and be that much stronger vs Wire
1:46 - Red Gem is worth more than lockpick so of course he steals that. A sufficiently strong Matthew could attack instead and get the Red Gem + more XP, but his Matthew is not leveled much.
1: 54 - Note his Matthew, it has +1 Def and +1 HP over baseline. This is critical. This means the archer kills him in 3 shots, not 2. He relies on that heavily.
1:59 - Wolf Beil will kill Wire in 2 hits (1 on this turn and then 1 on Wire's turn), allowing him to win the chapter now. Handaxe would be safer but take forever. However, he's relying entirely on luck here since Wire gets 2 chance to kill him. He's also relying on the fact that the boss goes earlier than the archer does, so the archer will not get a shot at Hector before Wire dies next turn and the level ends. Again, AI manipulation. Bosses typically go before everyone else and you can sometimes use that.
2:05 - He REALLY lucked out. That was like 5% chance of success in the fight with Wire alone I think.
So looking at the video, he had the following advantages that one can't always have:
1) Best possible outcome of fight vs first soldier. He needed to hit twice and be missed and that's what happened. With true hit the odds were about 38% of that outcome.
2) Matthew is tough enough to take 3 arrows. Generally speaking, any Matthew who has gained a single point of HP, let alone defense, can do this. A baseline Matthew cannot, so it's a good thing he used Matthew in Lyn's story. If Matthew isn't tough enough, Matthew needs to dodge one of the two attacks. Each has a better than 80% chance to hit.
3) Hector's handaxe attack on the archer MUST work or Matthew is doomed regardless. The odds of that are about 68%
4) Matthew must hit both times vs the archer. The odds are close to 99%. (Annoying one of my strats failed because Matthew missed. That bungler!)
5) Again, Hector had to double hit vs the second soldier. This time it didn't matter if the soldier hit or not. True odds about 90%
6) Hector must hit Wire twice and Wire must never hit Hector. Odds? 4%.
There were really pretty much no cases where bad luck hampered him. The chance of one more soldier hitting him is much higher than one more archer missing him. So multiplying out those numbers: the probability of success I calculate is <1% with this strategy. Assuming Matthew has leveled up and gained HP or Def already. Otherwise it will be <.5%
Some might argue that a strategy with a roughly 1% success rate is terrible, but in fact this was very clever- just not well executed. I kinda suspect some RNG manipulation to make it work too. In any case, the strategy involves several notably good choices: 1) take the top route. Time is the most importance thing here. 2) Do not open the door to the treasure room. This is very counterintuitive but very smart. He knew the thief would stupidly open that for him during the end of turn 5, so that freed Matthew to forge ahead. 3) Have Hector ready to charge once Matthew opens the locked door. If Hector does not go out full throttle immediately after that door is opened, he will probably be blocked in or killed. 4) ignore the remaining archer when choosing to attack Wire . This archer is a non-factor because either Wire or Hector will die before the AI considers using the archer.
Anyway, I'll show you what I did once I finish baking some cookies. What I really hope is to encourage you guys to think about what I do in the same kind of fashion I think about what this guy did. Notice when I screw up or when I rely on luck and think about what could be done differently. But also think about what things we did were smart and consider how you can use that thinking yourself. For example, I would never have ignored that archer on the last turn before watching this video. I would have assumed the AI would play smart and have Wire fight last. Realizing that the AI moves bosses first could come in handy elsewhere.