Mids' Reborn Unofficial (MRBU)

[Felis Noctu]

11 hours ago, Bopper said:

The Fly and Hover stuff is definitely weird in Mids. It's funny, if you look at the Info of Hover/Fly, the calculated FlySpd is wrong, but if you click the Effects tab, the FlySpd is correct. I can't tell exactly what is wrong with each, but it seems for Hover the FlySpd on the Info tab is not doing what it should be:

FlySpd = 50% * (1+Enh)+50%-101%

But instead I think it's doing this:

FlySpd = 50% * (1+Enh) + 50% + 50%*(-101%) = 50% * (1+Enh) - 0.5%

So instead of adding everything together, the last debuff is multiplying with the middle buff.

 

Then with Fly, it gets even weirder. It should be doing this:

FlySpd = 136.5% * (1+Enh) + 80% + 100% - 101%

But instead I THINK IT MIGHT be doing this:

FlySpd = 136.5% * (1+Enh) + 80% * (100%-101%) * (-101%) =  136.5% * (1+Enh) + 0.808%

 

I very likely am wrong on some of that. But looking at the numbers, it seemed like the only thing that fit. There seemed to be some interaction between the unenhanceable fly buffs/debuffs that seemed multiplicative instead of additive.

 

And I won't even get started with how the actual fly speed is calculated on the View Totals window. Those numbers are bonkers.

Yeah, you're definitely on the right track there. In essence it's something to do with how Mids builds its calculations, whether it's the operational order or that it isn't allowing itself to go below/affect certain values (like the base value) or what.

 

This is something I actually pointed out before, I forget if it was in the Mids discord or privately with Zed or what. It wasn't really work on fixing it so much as just scratching my head and making comparisons between in-game and Mids numbers to try and figure out where the math was going wrong, and the only conclusion I could come to was that in one way or another (like what you were doing) the flight speed debuff isn't being handled correctly.

 

I'm poking around at it again right now to see if I can figure out what's going on, so hopefully I'll have something to show for it in a bit.

[Bopper]

10 minutes ago, Felis Noctu said:

Yeah, you're definitely on the right track there. In essence it's something to do with how Mids builds its calculations, whether it's the operational order or that it isn't allowing itself to go below/affect certain values (like the base value) or what.

 

This is something I actually pointed out before, I forget if it was in the Mids discord or privately with Zed or what. It wasn't really work on fixing it so much as just scratching my head and making comparisons between in-game and Mids numbers to try and figure out where the math was going wrong, and the only conclusion I could come to was that in one way or another (like what you were doing) the flight speed debuff isn't being handled correctly.

 

I'm poking around at it again right now to see if I can figure out what's going on, so hopefully I'll have something to show for it in a bit.

Is there any human readable code you can share? I'm very curious where one calculation is made and is correct, then the same calculation is made and it's wrong, and where a different calculation is made and it doesnt follow any of the previous calculations. It's certainly a brain teaser.

[Felis Noctu]

5 minutes ago, Bopper said:

Is there any human readable code you can share? I'm very curious where one calculation is made and is correct, then the same calculation is made and it's wrong, and where a different calculation is made and it doesnt follow any of the previous calculations. It's certainly a brain teaser.

The official repo is here: https://github.com/Crytilis/mids-reborn-hero-designer

Our private changes haven't been published anywhere since we have no idea how stable they are, and they're mostly DB-editing convenience anyway. If/when we do release our own updates we'll stick the code up as well.

Edited March 28, 2020 by Felis Noctu

[Felis Noctu]

It's not Fly. It's not Hover.

 

It's Swift. Swift is causing the miscalculation.

 

I made a copy of Hover and removed all SpeedFly bonuses from it, just allowed it to enable flight mode, and checked the numbers again. Swift adds a full 8mph to flight speed when activated, despite only adding 13.65% SpeedFlying of 21.5mph (~2.9mph).

 

This and more news at 11.

 

EDIT:

More news below. It isn't Swift specifically either, Mids itself is somehow miscalculating the SpeedFlying bonus, and it applies to everything that grants flight speed.

Edited March 29, 2020 by Felis Noctu

[Caulderone]

Swift may be odd, but it can be unchecked.  I started a build and picked nothing but Hover, no other powers placed.  I put 3 fly in Hover, and unchecked Swift.

Result:

49.9MPH Hover?

 

Yeah, that isn't right.

[Bopper]

24 minutes ago, Felis Noctu said:

It's not Fly. It's not Hover.

 

It's Swift. Swift is causing the miscalculation.

 

I made a copy of Hover and removed all SpeedFly bonuses from it, just allowed it to enable flight mode, and checked the numbers again. Swift adds a full 8mph to flight speed when activated, despite only adding 13.65% SpeedFlying of 21.5mph (~2.9mph).

 

This and more news at 11.

I tested the same last night, and I did notice a huge jump when activating Swift, but the numbers were still wrong when it was off.

[Felis Noctu]

2 minutes ago, Bopper said:

I tested the same last night, and I did notice a huge jump when activating Swift, but the numbers were still wrong when it was off.

Well, Hover makes you lose 1% flight speed from your base, so you'll see a small decrease there. As for Fly I haven't actually compared numbers yet. I'm still trying to get Swift's values to a point where it's calculating similar to the game, that way I have a proper comparison.

Edited March 29, 2020 by Felis Noctu

[Bopper]

2 minutes ago, Felis Noctu said:

Well, Hover makes you lose 1% flight speed from your base, so you'll see a small decrease there. As for Fly I haven't actually compared numbers yet. I'm still trying to get Swift's values to a point where it's calculating similar to the game, that way I have a proper comparison.

Yeah, check out Fly's uncapped speed. It's silly large, like 158 mph.

 

I'll get on the computer now and see if there is anything consistent with numbers that might help narrow the search

Edited March 29, 2020 by Bopper

[Felis Noctu]

4 minutes ago, Bopper said:

Yeah, check out Fly's uncapped speed. It's silly large, like 158 mph.

 

I'll get on the computer now and see if there is anything consistent with numbers that might help narrow the search

Oh there's absolutely a problem with it, but I'm assuming the actual issue is related. Hover when tested by itself is practically unnoticeable because the Movement totals round to one decimal place, but I'm sure that 1% is wrong as well.

 

I asked Procat if he had the time to track down the rounding and open it up to at least 2 places for me, but I don't know when that'll happen, so I'm just doing my best for now.

[Felis Noctu]

I found the code for it and sent the changed line to him, he said about 20 minutes or so to finish up what he's doing. The display isn't the actual issue, but at least it gives us an idea of how far off things are.

[Bopper]

If it helps with troubleshooting, here is a chart I put together showing different settings with Fly/Hover and Swift, with different enhancement levels in each power (up to 2 slots of a level 50 IO in Fly and Hover, up to 1 slot of level 50 IO in Swift).  If I designate something as off, it means it is not active and is not contributing in any way to the totals. Columns F, G, and H (Info, Effects, View Totals) are directly from Mid's, which show the numbers with respect to Fly/Hover. Column G (The Effects Tab when looking at a power in Mids) is always calculating the FlySpd correctly. In Column J and K, I show what the total Fly buff and MPH should be (including Base). These are the numbers you would have in game under these conditions (assuming no rounding error). I haven't looked over the numbers enough (will later) but if someone wants a head start, maybe you see a consistency with the errors.

 

[Felis Noctu]

TL;DR:

I think this is going to require a source code change, not a database change. It seems it isn't the effect values that are the problem but how they fit into the calculations.

 

---

 

32 minutes ago, Bopper said:

[snip]

That does help immensely, thank you! After confirming my two screenshots with that, using that sheet is much easier.

 

So unfortunately Procat's too tired to be able to process it right now. He made the changes and now the Movement totals are rounding to 2 places instead of 1, so that helps. However, the only thing I can imagine going on here is it's something internally, not the database. Given all the testing I've done so far with all different configurations of powers, the only thing that's consistent in the equation is that the SpeedFlying bonus is not doing the math it should.

 

What usually throws me, at least at first, is Swift's value in-game. It's gotta be a display issue because all the other numbers are correct, but I always trip over it.

 

 

Anyway! Yeah, I think this is going to require a source code change, not a database change. It seems it isn't the effect values that are the problem but how they fit into the calculations.

Edited March 29, 2020 by Felis Noctu

[Bopper]

15 minutes ago, Felis Noctu said:

What usually throws me, at least at first, is Swift's value in-game. It's gotta be a display issue because all the other numbers are correct, but I always trip over it.

 

 

The only numbers that are correct in game (referring to the right side of the screen shot, the combat attributes) is the Flying Speed total and the Base (+21.48 mph). All the numbers below that (Swift, Hover, Hover, Hover) are not using the correct Base Fly Speed, they are using the regular base speed of 14.32 mph. If you multiply the numbers shown by 1.5, you will get the true values.

Edited March 29, 2020 by Bopper

[Felis Noctu]

Ah. Right. Hover. Because that 1% is practically unnoticeable so the end result is basically the same.

[Bopper]

2 minutes ago, Felis Noctu said:

Ah. Right. Hover. Because that 1% is practically unnoticeable so the end result is basically the same.

Yup. Throw in a fly speed enhancement into Hover and it becomes more noticable. Number 6 confirmed my findings awhile back, but said there was no simple way to fix the combat attributes display. They dont see the 1.5x multiplier Fly Speed gets

[Bopper]

I looked at the numbers some more and played with a few things, and I may have made a discovery. I added a new column (column I ), that shows the equivalent Total Fly % (including base) that compares what the bad View Total shows relative to Base Fly Speed. So column I is just column H divided by 21.48 mph (21.477... actually).

 

I noticed a trend with that, so I added a new column that just looked at the + Fly Boost (not including base). So columns M and N are simply columns I and J subtracted by 100%, just to isolate the Fly Speed bonus. Then I looked at that ratio (column O), and almost entirely the Fly Speed bonus used to calculate the Final Fly Speed in the View Totals window is 273%⁺ of what the actual fly speed bonus should be. 

 

If I were looking through source code, I would want to find where this consistent 273% strength to Fly Speed is coming from. I'll also take a look at Group Fly and any other fly powers in the game to see if the Fly Speed Total is also showing this +273%.

 

Edit: I took a look at the Base Fly Speed and the Fly Speed Cap. Guess what 58.63/21.48 = 2.73.  I don't know where things are going wrong, but this 273% (or 2.73x) seems to be tied in some way with the Base Fly Speed and the Fly Speed Cap.

 

⁺Coincidentally, the Jump Pack (and Steam Jump) in the game shows in its details of giving a +273% strength to Fly Speed. In game, it's actually 300%, but still, the 273% is quite the coincidence. 

 

Edited April 28, 2020 by Bopper
additional findings

[Bopper]

Checked out Group Fly and it follows the same trend. The +Fly Speed bonus is always multiplied by 2.73 when calculating View Total's Fly Speed. I updated the above chart with Group Fly and cleaned up my formulas to not introduce so much rounding error (I'm looking at you row 15). 

 

[Felis Noctu]

I really appreciate all the work you're putting into this! It's extremely helpful. Now we have a pattern to dig around in the source for.

 

I did notice while looking for the display code that there's a lot of strange movement-related things going on. Dev comments as well as multiple older versions of calculations commented out, more dense than most places I've skimmed previously. I'm guessing this isn't the first time people have played around in this area.

Edited March 29, 2020 by Felis Noctu

[Bopper]

1 hour ago, Felis Noctu said:

I really appreciate all the work you're putting into this! It's extremely helpful. Now we have a pattern to dig around in the source for.

 

I did notice while looking for the display code that there's a lot of strange movement-related things going on. Dev comments as well as multiple older versions of calculations commented out, more dense than most places I've skimmed previously. I'm guessing this isn't the first time people have played around in this area.

Wouldn't it be a trip if you find in the source code that it was scaling the Fly Speed Boosts to Max Fly Speed instead of Base Fly Speed? If only it's that simple :).

 

I'm always happy to help. This issue has puzzled me for some time, and I guess while in self-isolation I need a good puzzle to solve. If you can point me to the exact source code you're referring to, I can try to poke around and see what I can try to understand.

[Zed]

(reading the whole thread)

Whoever asked about an interface revamp: our top priority right now is to squash bugs and inconsistencies in the database. But it seems to me making cosmetic changes or dusting off the UI would be like re-building a skyscrapper from scratch, the same, but using rods of titanium instead of steel.

[Felis Noctu]

On 3/28/2020 at 9:25 AM, r0y said:

As always, great work keeping the database updated and as close to the game as humanly possible!

 

Is there any plans to completely re-vamp the interface?  I mean, ground-up type re-invention / re-imagining level stuff?  This (theoretically) can be done separate from the database / numbers stuff, by a separate team...?

 

Just curious.

5 hours ago, Zed said:

(reading the whole thread)

Whoever asked about an interface revamp: our top priority right now is to squash bugs and inconsistencies in the database. But it seems to me making cosmetic changes or dusting off the UI would be like re-building a skyscrapper from scratch, the same, but using rods of titanium instead of steel.

 

Yeah, redoing the interface probably isn't worth the effort involved to get a roughly similar result. A lot of it is tied to the functionality directly, so it'd be a lot more than just changing the visual forms involved.

  

 

---

(crosspost from official Mids thread)

 

7 hours ago, Redlynne said:

12 hours ago, Apparition said:

17 hours ago, Rikis said:

Currently tried 2.6.0.7 but I can't get over the all blue color theme. Before, when I clicked on Villain, it would switch to the better red color theme. Seems with 2.6, this doesn't work and I can't see how to change the colors myself in any of the options.

It was changed to blue to aid with color blindness, IIRC.  I prefer the red theme as well, but c'est la vie.

Kind of wondering why Blue/Red/Gold wasn't made an option to deal with the color blindness issue ...

 

The color scheme seems to be built primarily with images (really just one) and some color values for things like the text of selected powers, the summary box, etc. While doing an overhaul of the interface like mentioned above probably isn't worth the time, adding color schemes probably wouldn't be all that difficult. The only thing the Hero/Villain button seems to do at this point is change the names of Accolades and some Temp powers.

 

I'd say this is something we could probably add to the MRBU docket if @Procat is up for dealing with the code. Or at least the compiling if I can find the right lines. 😛 It just isn't going to happen until after we're satisfied with the state of the main functionality. I miss the villain theme myself, and adding a Praetorian golden theme would be neat.

[Felis Noctu]

13 hours ago, Bopper said:

Wouldn't it be a trip if you find in the source code that it was scaling the Fly Speed Boosts to Max Fly Speed instead of Base Fly Speed? If only it's that simple :).

Wouldn't it be a trip if you find in the source code that it was scaling the Fly Speed Boosts to Max Fly Speed instead of Base Fly Speed?

in the source code that it was scaling the Fly Speed Boosts to Max Fly Speed instead of Base Fly Speed?

it was scaling the Fly Speed Boosts to Max Fly Speed

Max Fly Speed

 

If only it's that simple :).

 

clsToonX.cs : Line 476 :

            Totals.FlySpd = Statistics.BaseFlySpeed + Math.Max(_selfBuffs.Effect[11], -0.9f) * Statistics.MaxFlySpeed;

(changed to)

            Totals.FlySpd = Statistics.BaseFlySpeed + Math.Max(_selfBuffs.Effect[11], -0.9f) * Statistics.BaseFlySpeed;

 

I was looking at it yesterday. Looking right at it. Didn't even process it at the time. So yeah, it's fixed!

 

 

Future releases of MRBU, as stated previously, will include at the very least an updated executable and related DLL file.

[Bopper]

4 minutes ago, Felis Noctu said:

Wouldn't it be a trip if you find in the source code that it was scaling the Fly Speed Boosts to Max Fly Speed instead of Base Fly Speed?

in the source code that it was scaling the Fly Speed Boosts to Max Fly Speed instead of Base Fly Speed?

it was scaling the Fly Speed Boosts to Max Fly Speed

Max Fly Speed

 

If only it's that simple :).

 

clsToonX.cs : Line 476 :

            Totals.FlySpd = Statistics.BaseFlySpeed + Math.Max(_selfBuffs.Effect[11], -0.9f) * Statistics.MaxFlySpeed;

(changed to)

            Totals.FlySpd = Statistics.BaseFlySpeed + Math.Max(_selfBuffs.Effect[11], -0.9f) * Statistics.BaseFlySpeed;

 

I was looking at it yesterday. Looking right at it. Didn't even process it at the time. So yeah, it's fixed!

 

 

Future releases of MRBU, as stated previously, will include at the very least an updated executable and related DLL file.

So awesome. It's funny how the simplest things cause the biggest problems. They are so easy to overlook. I'm glad this finally got resolved. Great work.

[plainguy]

Will this mess up my OLD builds I have already ? 

 

Just want to know where I stand before I try this as I don't want to lose what I have now. 

But I am sure I am missing out because I don't have all the current updates especially with the new set coming out soon.

[Caulderone]

30 minutes ago, plainguy said:

Will this mess up my OLD builds I have already ?

Nothing I made in 2.6.0.7 has been effected.  Things from pre-2.6 may be weird.