I support doing a main unit based Any%. That’s not a bad idea. As long as those restrictions of “main clan” are only for full game runs. With Any%, having that rule would be very restrictive and a lot less fun.

As for unit categories though, obviously TSA and DF, perhaps Shadowseige, but since you can’t start with him it’d be difficult to make a run with him. Same as Dante or other rare units. However, I could see maybe something for the units with Sweep, though without starting Quick (like TSA) they are at a disadvantage. Perhaps just a 3rd catchall category of Any% where DF and TSA are not allowed. I don’t know really.

With how uncommonly you'll see Dante in a run (especially if you skip Celestial Alcoves to save time), I don't think we'd need another category for him. After my Banished/Hellhorned record (6:12), I really think having a single Any% that allows any combo of clans and units would be ideal. If not that, my second vote would be how it was before with Any% allowing anything, but being divided by primary clan.

TSA might be the best overall (I don't have enough runs to prove one way or another), but Demon Fiend could just as easily beat him given the right tools. Even if TSA or DF come out to be clear winners, banning either of them seems a bit silly to me. There's always going to be a meta in speedrunning, and that's perfectly fine.

The idea of separate clan categories makes sense when setting up originally. However, now that I’ve run all the clans and watched several runs, it feels like the same run each time.

I’ve also submitted a Demon Fiend run that beats my best TSA run, utilizing some Melee Weakness from Banished. It’s pretty interesting thinking of a strategy that beats all other strategies.

With that thought, I would love a single separate category for C10T! I mainly play C10T random/random/random, but playing for speed would be a new twist.

Since the split of Any% TSA and Any% (banning TSA), I've done several Any% runs now (with a few successes). I've tried a new strategy (new for me): Demon Fiend with Pyre of Savagery, which is very similar to the TSA strat with the built in Sweep (using Trample instead) and auto-scaling with the Pyre. Here are my thoughts on the split:

Pros:

-It is a bit harder to keep DF alive compared to slapping Organ Harvester on TSA, which means each run requires finding 1 or 2 of a few different things to keep him alive. This does make runs a little bit unique to each other (though not too much in my experience).

-There's a little bit more thought and consideration to be had at Steel shops and Arms shops, since you're wanting Quick, Multistrike, and probably some way of keeping DF alive (another unit). There are a couple options for each (though Multistrike can be tougher), instead of literally just looking for 2x Multistrike on TSA and OH.

Cons:

-The runs still felt similar to a TSA run, without too much difference in average time (I don't have a huge pool of data to base that on though).

-I almost felt like I had to reset if Dominion is my final boss, as he requires so much armor to keep DF alive. I did get a good run with him, but there were several failed runs because of him. This adds one more reason to reset when finding a run start, which I don't really care for, but isn't a huge deal since it takes a second or 2.

-It feels like I'm just hoping to see Quick early, or else I'm going to die pretty quickly (and I'm wasting precious time watching the enemies attack me first).

-It still feels like a single strategy is going to dominate the leaderboards.

All that being said, I really don't know if having 2 Any% categories is necessary, nor do I actually think we need a category for each clan. With both Lazarus and Hellhorned, they essentially have everything you need to win within their own clan, with some clans offering a boost here and there. So it feels like every run, despite the clans used, will be the same thing with only slight differences.

Being able to see who can get the best overall time, no matter which clan or unit they use, would be my preference. I know that differs a lot from the MT1 leaderboard, but I don't see that as a bad thing, given how we have double the clans now. It might also make the leaderboard look a little less empty. However, that change isn't necessarily the main point of this post, just another thought when talking about Any% as a whole.

I'd be happy to hear other's thoughts on having separate leaderboards for each clan and on the split of TSA and no-TSA. Like I said, I don't have a huge amount of data yet, but these are just my experiences from running both types of Any% several times, and the cons seem to outweigh the pros by just a bit.

Honestly, I'm fine with just a single Any% category, for reasons I'll put below. Though I'm not sure if "Any% TSA" is the right way to do it. While TSA certainly seems like the fastest, and I can't theorize anything else better, there could be something else that's found. It also seems odd for a speedrunning group to say, "The fastest strategy is it's own category, and the second fastest will be their own category." Especially when it's essentially the same run. It's not like a skip or glitch is found that makes it a brand new category or anything. We'll end up with a very similar problem (though not quite as easy to pull off) with Demon Fiend.

That being said, after doing all 10 clans, just sending TSA the whole time, it felt like having the 10 clans as separate categories was a little repetitive. It also makes it difficult to see what the actual fastest time for the game is, and track that history.

I definitely can see where it feels too restrictive making the player use each clan once, and perhaps there's a middle ground somewhere. But after doing the 10 clans, I'm turned off from wanting to run Full Game just to do 10 back-to-back TSA runs. Thinking about what clan combos would work well to have fast and reliable runs though, that sounds fairly interesting.

Also, I apologize if it seems like I'm coming in as a new runner and just wanting to change everything up. I'm definitely not wanting to go against the grain or devalue any runs. I simply wanted to give my thoughts after doing several runs and see if others felt the same way. If the consensus is to keep everything the same, I'm perfectly fine with that.

To the macros: that makes perfect sense. The only reason it was even a consideration for the Full Game was your comment about adding a set time for each reset. But it sounds like I misunderstood what you meant.

I did have a couple thoughts on this run. Currently as it stands, it seems like it'll be 5 (or 10) less-than-ideal-single runs, all bunched together. I say less than ideal, as you can't hope for the absolute best like Any%, and you really need your runs to stay alive, so you'll play a little more careful. This is all a presumption based on how I'm planning it out for myself. It would likely be all Hellhorned or Lazarus as a primary or secondary clan (Hellhorned would be more durable, but Lazarus is faster with TSA), same pyre every time, and basically the same thing over and over. This honestly seems boring to play and watch.

I watched a MT1 Full Game run and saw that there's a rule stating:

You MUST use each clan once as a primary and once as a secondary during your runs.

I wonder if this would be something worth considering for Full Game to bring a bit more strategy and thought into it. This would certainly change it from, "same strategy every time", to "What combos will work well together and what pieces am I looking for on THIS run?" I know there's a lot more clans to work with in MT2, and i'm not sure how it would work with the split Full Game (only MT1 or only MT2 clans), but perhaps it could be tweaked to say:

"You may only use each clan once as a primary and once as a secondary during your runs."

For the split Full Game runs, that'd allow for mixing and matching between MT1/2 clans but still only allow a clan once per run. And for the full Full Game (all 10), that'd be interesting to see what would be the best combos to use.

If this idea is considered, I think we should take into account Wurmkin and Railforged as either allowing them (kind of substitutes for less than ideal clans), or banning them (since not everyone will have the paid-DLC for Railforged, and consistency and fairness may be a factor).

I'm curious how you and others feel about the use of a macro for restarting runs, mainly for the Full Game category. Both my Steam Deck and laptop have a macro I've created where I can press a single button and open the menu followed by pressing the reset run button. I only use this for getting a new run started, saving some wrist pain on my laptop and saving effort on the controller.

In a single run, this macro does not effect my time or run as it's only used for starting the run, not to gain an advantage within the run. However, I see that it could be seen as cheating when doing a Full Game run (whenever I get to that point), given what we discussed about restarting runs being part of the time. That being said, if we want to go with your suggestion of

"So, for every reset, I could add on the time displayed (probably usually about 3-5s I would assume) and then add that on at the end of the clan's run, and state how much time was added per legion in the comment section under the run."

I think we could say every reset adds X time to that clan's run and whether someone uses a macro to reset in 1 second or manually does it in 2 seconds+, it will always add X time per reset. That X should be considered based on load times and reasonable time to reset by players, but that's another discussion.

Also, if there's something I'm not seeing about using my macros for starting single runs, please let me know. I certainly don't want to gain an unfair advantage, however I also don't like a single repetitive motion when it's not a part of the speedrun itself.

I think the first option would be the best. There would be some incentive to reset to get a decent start for each run but also a drive to just push forward to not waste too much time. It's also interesting to see what a player can do with what they're given, and just trying their best, even with subpar starting cards. I'd be curious about how much time it's worth reseting for a good unit, due to how much time will be lost by not having it.

I know I haven't submitted any runs here, so my opinion may not have as much weight, but I did 100% MT1 and 2 and love the games a lot. I've also enjoyed speedrunning (both watching and participating).

I know for Any%, reseting for a good start is ideal. However, would that be allowed for a Full Game speedrun? Let's say I start run 1, quick-reset for a good start, finish, and show IGT. Then, I start the next clan's run and reset until I get a good run there, and show IGT, etc.? Would the reseting count towards the total time?

Or would I need to reset the first clan's run (for an ideal start of course) and then once I've got a good run there, I just have to power through with whatever the game gives me on each subsequent run for the other clans, no reseting (unless I lose, but that's counted towards my total time)?

I'm just trying to get a clear picture of how a run like this would look, and unfortunately there's no video to watch and the rules are a little ambiguous.

So will this rule change invalidate any runs that were previously done on an emulator that is now banned?

Another idea I just had was an average of all your runs. It might be more enjoyable to do a casual run against your averages, rather than against your best time. Just add up all the splits for each individual segment, and average it based on the number of times that are saved for that segment. I'm sure there's a way to save the total, without having to save ever single run's times, as that could take up some memory after a while, but I'm not a programming expert. You should also make it so if you decide not to save the time when you reset, it won't be a part of the average, but that's probably obvious.

The way you explained the splitception is cool, but I will say that the 2nd way of doing it just sounds confusing to the runner. I doubt I would want to use it myself. I can see myself using splitception for 14 Tali, 40 Orb, and probably Any%, just because they don't do revisits, but just using the normal splits for 100%.

As for the counter idea, the only thing I can think of would be counting gems. Like for 14 Tali, you need a certain number of gems at a few different spots in the run, so gem management is crucial. Maybe have it so that you can put in the splits (just like the time for the level) the gems you want to have at the end of each level. Then during the run, when you split (if the user has checked a box saying this split needs gems) a dialog box will come up, and you enter in the gems you have. It will show the difference, whether you're high or low, so you know how many extra you need, or how many superfluous gems you can skip. You could also use it to count dragons, orbs, or eggs respectively, just have a different key press to add one to that counter. I don't know how necessary either of those are, or if I would really need them, but it might be a nice feature to others.

First of all, I would greatly appreciate this, as my timer, Llanfair, likes to slow down my emulator, and I'm looking for a new one. I'd also be willing to playtest any beta versions for Mac.

Some things I would love to have would be: -Resizing the window -Customize font size of splits/timer/title/etc. -Custom colors for splits (saving time, gaining time, glod) -Option for icons for splits -Ability to scroll through your splits, if there are a lot of them (such as 100% runs)

I like the idea of splitception, but with runs like 100% that backtrack and do levels way out of order, how would that work?

Update: The newest iso I got is fine (I'm not sure about the one you linked me to, but I would guess it's fine), but PCSX apparently doesn't have the plugins, or whatever is needed, to get past the anti-piracy things in Spyro 3. I'm having to run psxfin (per the advice of a friend) through a program that let's Mac run .exe programs, but it's working. Thanks for trying to help, I do appreciate it.

The v1.1 link you gave me didn't work, and neither did another iso I found which claimed to fix these issues (no-paradox). Really not sure where to go from here.

I use PCSX, as that's the only one that's worked on Mac for me so far. I'll try the v1.1 when I have some time, and report back if that's fixed it.

So I've been learning 100 egg (thanks to TOTO's guide), and my emulator/rom won't let me fight the Sorceress. I go in the fight, and after about 30 seconds, it warps be into the first hub, with no gems and no eggs. I've also experienced something where I die to Spike or Scorch, and after beating them, can no longer get a better Sparx than green or blue. I know Insomniac put anti-piracy things in the game, but I also see that people play this on emulator successfully. Is there a fix to this?

I'm also running the 1.0 version, not the greatest hits (1.1) version, if that matters..

I was messing around with the ones at the start, but couldn't get a high enough proxy to skip the whole level (maybe with TAS it'd be possible). Athough you could potentially get a high enough proxy to get over to the lower platform (two vases and a snow wizard), allowing you to easily setup for Saboom's wall glide Moneybags skip. But like you said, snowballs are hard to proxy, and just doing this video with savestates took about 10-15 minutes.

I was playing around with the fodder in Crystal Glacier, and found you could get some good proxies off of them. After some messing around with it, I found what could be a new strat for getting into the talisman room:

As the video says, you do gain ~7-8 seconds, but you lose about 25 or 30 gems. Could the time save make up for the gem loss?