PlanetaryGenocide reacted to 8_Hussars in Analysis of 9853 Replays
Please no bully, its a work in progress.
I finally buckled down and made the WoT replay analyzer work for me and I managed to analyze 9853 of my matches I had available over the last few years (2018 to 2020-Q3). The Google Sheets link is below if you are interested (with all the raw data). I really hoped to gather Battle Tier Data and Number of arty but that is not available or very difficult to dig out. I figured out how to collect Battle Tier, MM schema (e.g. 3/5/7) and arty data. It is now presented as well.
The original intent was to examine the data in light of @Private_Miros data presented on EU forums and cross posted here (IIRC) and have something to counter all the "snowballs are increasing" and "match lengths under two minutes" deniers on the Official forums.
Not surprisingly they petty much confirmed @Private_Miros data. The Match Length numbers although have a few more anomalies on the low side and its NA server.
Matches Analyzed: 9853
Average Match Length: 00:07:06
Shortest Match Length: 00:01:57 (surprisingly this was a T10 Erlenburg Encounter and not Mines Encounter)
#Matches under 3 Minutes: 19 0.2%
#Matches under 4 Minutes: 263 2.7%
#Matches under 5 Minutes: 1344 13.6%
#Matches - "All vehicles destroyed" is Closer than 6 Tanks: 1792 18.2%
#Matches - "All vehicles destroyed" Score is 6 to 10 tanks: 4851 49.2%
#Matches - "All vehicles destroyed" Score is Greater than 10 tanks: 1837 18.6%
Game mode (Standard 75%, Encounter 15%, Assault 10%) appeared in generally accepted nominal proportions
Win Condition (All vehicles destroyed 85%, Base Captured 13%, Battle time expired 2%) appeared in generally accepted nominal proportions
Game Mode (Standard, Encounter, Assault) was not a significant factor.
Win Condition (All vehicles destroyed, Base Captured) was not a significant factor. (Battle time expired notwithstanding)
Match Results (Victory/Defeat) was not a significant factor.
Average Match Time was pretty much linear (and decreasing) from a score of 15:14 to 15:00
Tank Tier (NOT Battle Tier) was not a significant factor for Tiers 7+. Although match lengths shortened up by 30 seconds or so for Tiers 5 and 6. Suspect this is map pool related (see below) and not play related.
the three smallest maps (Ensk, Himmelsdorf, and Widepark) played faster by roughly -1:30 (about 00:05:30).
the larger maps (Prok, Mali, Erlenburg, Kharkov, and Fish Bay) played slower by roughly +1:00 (about 00:08:00). Standard Prok almost broke 9 minutes!
This crushed my soul...
PlanetaryGenocide reacted to Caomhanach in T49 -- Land Arty in a Bite-Size Package
Since I'm bad at tanks, and can't hit the broad side of a barn, I'm interested in how the new handling equipment pans out, especially since IRM has been mentioned a lot on equipment threads. And since I'm such a huge fucking nerd, I mathed it out. I used the formulas and info from
I used the stats from tanks.gg, and only used enhanced equipment if that option was available.
First, I tested out one gun handling option at a time.
So, unless I screwed up the math royally (very possible, I didn't say I was a smart nerd), the most surprising takeaway is how underwhelming the IRM is, and how surprisingly good the eGLD is compared to everything but the eVS. The cause is two-fold. First, the T49's long aimtime make the GLD significantly better compared to it's normal gun peers, and second, the IRM is seriously hurt by large hull and turret traverse speeds and dispersion penalties. This was confirmed for me after I initially did the calcs without accounting for the additional 10% speed. In that situation, eGLD didn't catch up to the IRM until about 2.6 seconds instead of about 1.4. That's a huge difference that just a little speed makes, given the T49's dispersion penalties.
Now, since it's not surprising that the eVS won huge, I then did the calcs with eVS plus one other handling item.
I'm becoming even more suspicious of my math at this point, since the IRM becomes even less viable here compared to eGLD. I suspect that the T49's relatively fast traverse speeds are playing a bigger role here than I initially thought. Again, I initially did the calcs without speed increase accounted for, and the eGLD caught up at again about 2.6 seconds, instead of about 0.8. Also, IRM was better than IAU at every point except for a fully aimed shot, as opposed to here with the updated calcs, where IAU is always better. I'll have to review my calcs again later, but I've checked my work several times so far, and it seems good. Also, I should probably include vents. The other interesting takeaway here is that while eVS + eGLD is generally the best option, it loses to eVS + IAU for both snap shooting AND fully zoomed accuracy.
And, for shits and giggles, I also did the options for three gun handling pieces, because why not?
The eVS + eGLD + IAU is the clear winner here.
Thoughts? Keep in mind, the T49 is unique, and probably penalizes IRM usage more than most other vehicles in the game, so these graphs can't be translated to general performance of these equipment pieces on other vehicles.
PlanetaryGenocide reacted to Wanderjar in T49 -- Land Arty in a Bite-Size Package
other way around unless you're sniping
Aim time is:
original radius * (1/e)^(time passed/aim time) or 36.39% per tick of the previous value not max value vs final value (as confirmed by WG devs)
but the IAU just lets the final accy be tighter and doesnt speed up aim. The only time the IAU is valid is when you get below your nominal final accuracy.
GLD/VStab both work ALL THE TIME when moving or being still/after firing penalties
IRM works by limiting the dispersion put into your gun equation, thus making your Vstab THAT much more efficient and quick.to aim in. so for a t49 Derp IRM+Vstab+CVS is probably your go-to solution unless you like sniping in it in which case the IAU has a valid case to be used.
I haven't had the time to look at the post 1.10 patch aiming math yet but I THINK the way the IRM is plumbed in is:
ComputedDisp = GunAccuracy * SQRT ((1 + DispMovingTraverse^2 + DispHullTraverse^2 + DispTurretTraverse^2) / 1.1 )
Full disclosure: I literally dont know yet. its just a guess knowing how WG maths are coded in
PlanetaryGenocide reacted to sr360 in T49 -- Land Arty in a Bite-Size Package
Oh EGLD is certainly better than anything if you sit and aim down on large bloom/long aim time tanks, but given that EGLD stops working the moment you adjust turret aim or move a bit, and that your calculations are predicated on fully aiming down (which many of us rarely do!), in practice, EGLD suffers. In practice, a lot of T49 shots are drive-by pens, circling and getting behind and derping someone's rear, with the occasional half aimed shot where you can sit with locked tracks and turret for a full aimdown. In most practical scenarios, having the slightly tighter circle from the IVS/IRM combo is better. No doubt there are those whose playstyle favors the GLD setup -- and some have immense success with it, such as Muscles1 (who runs a fully gun-centric setup). For someone like myself who found it hard to hit shots and found few opportunities to get a full aim down, the IRM+IVS combination has made a significant improvement to my T49 comfort.
What I find intriguing in your math is the IAU>IRM argument. If your math is correct, for snapshots the IAU+IVS combo is superior to the IRM+IVS combo?
PlanetaryGenocide reacted to sr360 in T49 -- Land Arty in a Bite-Size Package
This is a topic of some debate and some personal preference.
Improved VS is uncontested. Then either improved optics or improved vents. The third slot can either get optics, improved rotation, or commander’s vision system. @CraBeatOff runs iVents/IVS and rotates the third slot based on platoon (one vision centered tank and one handling centered). I can’t shoot as well so I run rotation/IVS for max gun handling buff with iOptics for the phat scouting. I thought about iVents for even better gun handling but the VR loss was too much.
PlanetaryGenocide reacted to hazzgar in [Sandbox] Equipment 2.0
Since the bonuses are bigger than what they are now some tanks will be even stronger. Faster games. Idiots at wg. They want us to farm credits for new equipment.
Igoring foliage camo or making it double is idiotic. That would literally mean spotting is broken as shit. Vents + Optics + Ignore camo on an elc even or t100 or ebr and everyone is fucked.
PlanetaryGenocide reacted to dustygator in [Sandbox] Equipment 2.0
Perfect should not be the enemy of good, yadda yadda.
An equipment overhaul has been needed for a long time (it's one of the things that has barely changed since beta times) and this looks to be a step in the right direction.
Limiting certain equipment by classes is a bit annoying. Binocs limited to only TDs but then preventing them from using optics could indirectly nerf some more mobile TDs like E25, etc. Category limits/bonuses aren't great either. It pigeonholes tanks into certain roles and can make them less flexible/fun to play. Also I could totally see categories leading to a new meta forming eg. skilled players generally prefer the Obj 277 to IS-7. That gap would be even bigger if the Obj 277 gets firepower + mobility while the IS-7 gets armor + mobility. strongest would probably be heavy tanks with two firepower categories or medium tanks with firepower + scouting. Splitting the vert stab into two pieces of equipment is a reasonable nerf, as I would arguable it's the most valuable current equipment (even more than rammer). Although I think it would've more sense to group hull movement + traverse vs turret (rather than hull movement + turret vs traverse). The current combination and benefit levels make vert stab (+20% on two sources) clearly stronger than improved rotation mechanism (+15% on one source) for all turreted vehicles. The additional forward/reverse speed says MPH in one section, % in another. The former would be much more useful; it's generally slower tanks that need more mobility help and 105% of 15kph is still really fucking slow. Scouting equipment is the most interesting. Jamming device (-2/4s spotted duration) seems useful for active scouts. Commander's Vision (-10/20% to camo on bushed vehicles) seems interesting but too low; most non-tiny bushes are ~35-50% camo. I would double it, or make it ignore foliage camo completely. Would need to break out a spreadsheet to figure out how panoramic triplex compares to optics. My guess is that it is net worse (more effective vs moving vehicles but not by enough to offset lower flexibility).
PlanetaryGenocide reacted to kolni in Tier 10 Struggles (response to Snoregasm)
As of now, every viable tier 10 tank either has some sort of gimmick to leverage or just an overtuned kit in general. The gimmicky ones usually have flaws to play around when facing them, but the ones with overtuned kits tend to simply roll you over as a lower tier, and you rarely get to pressure an advantage in full tier 10 games as the terms seems are generally unbalanced and hard to estimate. Hull down snipefights is one of the most common aspects of the game and it is absolutely terrible gameplay, just an example but you get the point.
Some examples of the first would be:
Autoloaders High alpha guns Super high mobility vehicles These are fairly simple to play around because you know what they are going to do every game. Niche vehicles tend to go into set plays per map much more often than not and as such they get much easier to predict and counter. EBR’s really suck but I’m simply going to flat out tell you that if you can’t hit them reliably, your aim is bad. Their wheels are complete BS game balance wise but at this point you should be able to adjust to it. It’s part of the game now so the thing you can do instead of whining (it is valitaded, I know) you should be focusing on what you can improve, in the EBR case that would simply be aiming. It’s a terrible introduction to the game but don’t let that stagnate your progress.
On the other hand we have tanks like Chieftain, 907, 430U and such that are just statistically overpowered vehicles in most metrics. They are never easy to fight and you’re going to have to learn that the hard way. Even the dumbest monkey in a Chieftain has a real chance to win duels against considerably better players because of his tank selection. Which leads me to the topic at hand: Tank selection is everything in tier 10. Playing off-meta means you directly disadvantage yourself against everyone playing meta before you even click battle. If you want to learn how tier 10 works in its current state you need to spend time with it, and you generally want to be in control and figure out what works and what doesn’t. How do you do that? Pick strong vehicles. You want to develop that decision making ability but you won’t get to make many decisions when you constantly have to let the enemy make its move because you have a worse tank.
Since the full tier 10 MM is by far the most common, this gets more important. You don’t have anything to free farm to inflate your DPG, you have to work for it. This means no auto-piloting whatsoever because there are too many guns that have big enough impact to ruin your game entirely of one mistake. Aim for non-losing gameplay rather than winning. I don’t find the meta particularly different, just the tank balance being off. I generally try to avoid the stuff I don’t want to fight until I have to, even if that means giving up where I initially wanted to go during the countdown. If you care about performance in games then tank selection is going to have to be a part of it.
I find the meta to be increasingly faster and faster, compared to 4-5 years ago where playing full games at 500 meters was a legitimate viable strategy for most mediums. In that sense, yes, there’s more aggression today than before because people are fighting not only against their enemies but teammates for performance. This means people are going to cut corners and find ways to do this more efficiently, which leads to higher tempo gameplay.
When do you play aggressively? When you have all the information needed to make that play. How likely is it that this play is going to work? Estimate it, try it, re-estimate the value of the play in said situation.
I am mostly a passive player, so my play is almost exclusively reactive. I deploy to positions I know well, damage to farm and most importantly have a fall back option. If I’m wrong I don’t want my game to end so I never purposefully do plays that are questionable but “would be good if they paid off”. That type of “what if” thinking needs to work the other way too. What if this goes wrong? What should I do? Have an exit strategy ready so your game doesn’t end because of a misread earlier on.
I’d say the meta is fairly different too, arty being the main difference as there’s basically always 3 of them. Not much you can do about that sadly, once an arty has decided that he’s going to kill you there’s basically nothing you can do to prevent him from singling you out. The advice I have there is to go dark just before arty reloads so that he just might switch to another target.
It’s a harder game than any other tiers for sure, but you also learn the most from it. If I go back to tier 9 again I’m fairly sure I’d just roll over everybody because I’m almost able to do that in tier 10 too. Tier 9 doesn’t have so many gimmick tanks to be constantly be aware of, and has less variables which makes it an easier tier to play. This is the place you want to be at, comfort means a lot and you should be playing where you are performing the best, perfecting your gameplay in a more comfortable setting makes it easier to take those with you and apply them to more nuanced concepts and situations. Keep doing what works, try and up the difficulty every now and then and see if your skill set holds up. If it does, great; if not then you can just go back and practice more.
The meta itself of tier 10 isn’t very different, people still play the same positions and the game flow works the same, just at higher stakes. If it gets too complicated you simply have to accept that you’re not quite there yet. Self-criticism is the biggest part of improvement and the more you do it the higher chances are you’ll end up a better player afterwards. Blaming external factors doesn’t help you. Arty sucks, EBRs suck, unbalanced maps etc etc. You can’t do anything to change those, so focus on playing around them. What can I do to simplify these problems for myself?
For pure aggressive play I’d recommend starting out small. High probability actions that still are dictating actions. Try to control the engagements, think out ways to achieve whatever it is you actually want to do, before you do them. An enemy tank needs to die that’s hard to kill? Find out ways to do it safely, and if you can’t figure one out then you have the options of gambling or thinking of something else. Take the latter option. You can still control areas without putting stakes into it. Don’t know how? Try something. I will always recommend safer play when stakes are high, but when you really are at a loss it’s not like you’re going to end up with a great idea anyway, so try any idea. Sometimes they work out, great. You didn’t learn shit but at least you saved a game that should have been bad.
I also recommend studying the minimap a lot, especially at decisive points of action. See a push happening? Then you want to know how it develops from there, is it common? Can you use this experience for something useful, like assuming it normally happens (after it has happened enough times). This lets you skip steps both in decision making and focus, letting you be more concerned with what’s happening on your screen rather than the map once you understand it. You need to understand the map before you can start making aggressive plays reliably, because most players react the same way to the same things. Knowing that, you know what said aggressively play will result in, how the enemy will react and now you can build upon that knowledge to think even further ahead of that aggressive play. More foresight leads to much more stable gameplay, so the biggest piece of advice to your questions (even though I didn’t really answer any of them in concrete terms) is that you should understand other players better. All the 48-52% players are bunched together anyway, they don’t try to get into other players’ heads at all and react exclusively to what they see happening on their screen and not the minimap. No foresight whatsoever, meaning you can abuse the living shit out of it and simply win over them by playing big picture. Similar to GMs beating mathematically perfect playing computers in chess. Or the macro concept in DoTA/League. It’s a super advanced concept that doesn’t really apply until you are comfortable enough mechanically to beat anyone you come across. You want to be at that point mechanically, because mechanics don’t require thinking once you are comfortable enough with your game knowledge to keep your head in big picture, all the time. You shouldn’t be considering things like “Where should I aim to pen this tank?” or “How much do I need to lead my shell?”; those things should be etched in your skin and happen automatically, because then you can start playing the map which is what tier 10 is all about. You don’t have any substantial mechanical advantages in tier 10, so your theoretical advantages are what you’re going to want to leverage. A good place to start would be staying at least 1 step ahead of what you’re currently doing, while simultaneously having a fallback if it didn’t work out. Try sticking to that rule at every point of the game, and if you find yourself drifting away from it, you failed. Try again. Once that becomes routine you can add more variables into the mix, like managing HP in ways to prevent pub monkeys yoloing you from actually succeeding, and the more questionable parts of what in reality should be possible know and not. The point is that if you are unaware of something you can’t prepare against it, meaning you have stopped being in control of your gameplay. Try and stop letting that happen and you will find yourself having much more stable games.
This is all in very broad terms, but concepts like this are hard to verbalize unless you make them super specific which usually only applies to the specific situations and little to actually draw from.
I don’t really feel that I properly answered your questions, more guiding you in the right direction of becoming comfortable with how tier 10 works. The hard truth is that there’s no easy fix to somehow getting better, it’s a grind that takes time and constantly humbling yourself enough to accept that even if the result of something going bad wasn’t directly your fault, you can’t dwell on that since it doesn’t improve you as a player. It happened, so it was a bad decision. Could you have known that? Try and figure those things out and start playing the game ahead of time. A good rule of thumb that has been working for me is always staying 30 seconds ahead of the map and checking if my assumptions are right, and why they were/weren’t. You need to get into the heads of other players and pull mind-gamey shit to be the protagonist player of every game, which is what everyone wants. When you are in control you can do whatever you want, but at a loss of what to do, take the backseat and let the game develop more until you start to understand what’s happening. This doesn’t mean doing nothing, just simpler gameplay where you have teammates to play around and good amounts of information to make the easier assumptions.
TL;DR - Take the blame for mistakes, even if you feel they weren’t yours. What could you have done to improve the situation, regardless of what actually happened? Do this every game and you’ll be improving. Tier 10 is all about eliminating mistakes rather than improving positives.
PlanetaryGenocide got a reaction from Archaic_One in Sandbox: HE Shells "Rebalancing"
Getting booty blasted by a T49 HE round is generally a lot less RNG-dependent than getting slapped by skycancer for 500 a pop by a shell that landed 3 meters away, IMO, but I do understand that it's not exactly the best feeling in the world either way.
That being said, removing an entire sub-class of tanks and playstyle is not the answer here
PlanetaryGenocide reacted to Diriz0n in Supertest: bat-chatillon 12t mle 54
I wish I could like Emil51 more, I really do. But it has significant drawbacks. Despite having better mobility than Emil, they didn't provide it Emil top speed. Didn't provide it Emil LFP either, it takes damage way easier. The acc + disp are pretty sad face compared to Emil, and its down 15O DPM too....No idea why it wasn't buffed in relation to Swede heavy tank re-work. In the tier....getting side armour over-matched by 12O+ sucks pretty bad, and the APCR pen just makes it seem dreadful at times.
Bump up the R0F to 5.6, give it 125 LFP, increase the accuracy to .38 and dispersion to .24/.24 and I think its solid after
PlanetaryGenocide reacted to Diriz0n in 59-16 | An overlooked gem?
Remember that Type 64 stats have to continually contend with pre-LT changes, where it met higher tiers. People nowdays have them with very high DPG like 1.5K, which just wasn't common at all seeing tier 9, or even Xs before that. Now, it is extremely strong, though it per se didn't receive buffs, everything around it was significantly nerfed.
59-16 has a powerful gun, for a light tank. Like AMX12T, it is quite ahead in firepower compared to T21, VK, MT25, T37. And unlike 12T, is worlds ahead in DPM. The problems with this thing, are low view range compared to VK or T21, and the very low shell velocity. Shell velocity even with APCR is mediocre
PlanetaryGenocide reacted to Diriz0n in Dawn of Industry power-rankings
okay, a tad about some history first. My first CW reward was IS5, with Fusion. I got T95E6 with Thugz, as my first T X Reward. Selected 121B as my second, a part of Supernova. Selected obj907 for my third, being a part of Pingu. Lastly, just purchased M60 with bonds from the store. I am debating whether or not I want to take part in D of I, as I won't be able to take part in all of it. I leave for work on the 23rd, nor am I in a super duper hipper turbo clan, so I'd be stuck. The remaining driving factors to sway me, are of course bonds and being able to assemble either Kreslavsky or Chieftain. But enough of that, every campaign I have been part of had different directional meta-game tank usage, so I'm just out to predict D of I.
Star-Road, God tier: obj 260, Chieftain, 279E, obj 430U, obj 907, EBR105, GC
Every tank here is self explanatory.
Gold cup, Tier 1: Super Conq, WZ5A, obj 277, IS7, VK72, bat chat, UDES, obj 140, T-100LT, Strv, T110E3, obj 268V4, M53/M55
Pretty much the strong tanks acting as filler and subs in tank locking, or clans without lineups of CW rewards. Nevertheless still good stuff. VK72 the new replacement for derper Type 5, with the added attacking ability because of 35Omm HEAT. TDs seem to continue to fall and fall in usage because speed is so important, so the good ones appear here. Russian light is a great backup, and works. M53/M55 though a great arty, just can't seem to deal enough damage to be God tiered
Silver cup, Tier 2: Kran, T57, Maus, T110E4, K-91, Leo, AMX13105, obj 261
Mediocre stuff that is passable. or extremely situational, like Maus for sponging or T57 for close range DPM. Kran doesn't have suitable HEAT, so its here. E4 for those close range maps where people want to use it as pseudo-heavy, and it can still do the old Type 5 dirty ass HE'ing. Sniper schnitzel pseudo-TD meds as well. Thought about placing TVP here, but does anyone TVP anymore? STB1 despite the buffs, is just what I think is an UDES that sucks. obj 261 can still work as arty, though the low ammo count and damage, oooofffff
Everything else is trash.
121B could have been good with its buffs, but the DPM and top speed are just too low. T62A is a bad 140, itself a poorer 907. 113 is a bad 430U, lol at the people that thought it would rival WZ. T110E5 is a bad Conqueror. AMX30 a bad Leo. 705 a bad VK72. 60TP is over rated, I'd never take this thing over VK72 or E4. Don't think progetto or TVP have what it takes for CW, though excellent in pubs. Type 5 is trash now and rightfully so, WG best decision ever. Badger sucks, really did it ever amount to anything with all its prior hype? Panzer wagon used to do exotic route blocking strategies on Mines/Karelia hill, but with EBR that doesn't work anymore. Feel free to chime in if I left anything out. And before people want to cry foul about hulldown STB1 or M48A2 or Cent AX - believe me UDES replaces all of them
Your autism has been proofread and fixed. -H