Leader deaths much too high in PvP grand campaign games

[Beethoven1]

Multiple grand campaign games which have lasted until 1943+ show that leader deaths are too high.

I will post this in multiple posts due to screenshot limits etc.

I will post from two example games from which I have gotten players to share the leader deaths with me (both of which have at least short summary forum AARs):


Game 1: David (Axis) vs Caedus (Soviets) ---- forum AAR: https://www.matrixgames.com/forums/view ... 2&t=402253

By the end of the game in late 1943, there were 62 Axis leader deaths.

leader3.png (10.16 KiB) Viewed 1412 times

This screenshot shows who the dead leaders are, although even this is too small to show all of them:

leader4.png (1.45 MiB) Viewed 1412 times

The important point though is that you can just read through the list of dead leaders and you can immediately tell that this is completely absurd. Huge numbers of these leaders are leaders which you will have heard of and which either are the best German leaders or else start off the game assigned to a corps HQ:

Ferdinand Schoerner
Josef Harpe
Walter Hartmann
Heinrich von Vietinghoff
Eberhard von Mackensen
Erich Brandenberger
Walter Weiss
Georg-Hans Reinhart
Karl Adolf Hollidt
Hans Hube
Erwin Jaenecke
Maximilian Fretter-Pico
Johannes Friessner
Hans Gollnick
Walter Model

These are only some of the leaders in the screenshot who I recognized (most of which have particularly high mech or in some cases infantry ratings). Keep in mind this does not even include leaders who are simply too far down the list to visually appear in the screenshot.

It is so extreme that it is a challenge to think of good German leaders who did NOT die on that list.

BTW, if you follow the link to the AAR, this is a good relevant quote from Caedus from the AAR:

3. Pockets. Due to their ability to win single battles easily, soviets have a lot of units and movement points around to flood the gaps. At the same time, soviets can move a lot of units behind the front to prevent meningful breakthroughs by sheer mass. This lead to lots of german units surrendering, as well as an insane amount of leader kills, deminishing the strenght of an army (or corps) for eternity. I had the feeling, that leader deaths are way too likely when HQs are getting displaced.

I am aware that the game is attempting to try to match some sort of historical leader death statistics, and perhaps this does in fact roughly match those desired statistics in AI vs AI games, or in single player games with an Axis player against Soviet AI. However, this is not the case in multiplayer games at all, and the situation is very different.

And importantly, as far as I am aware, the statistics on leader deaths that the game is trying to match were historically mostly deaths of divisional or lower level generals (not corps leaders, army leaders, or army group leaders). But every single one of the leaders who dies in the game is a corps commander or higher!!!!!!!!!!!!!! This leads to an absurd massacre of the highest levels of the elite high ranking German officer corps by the end of many games.

[Beethoven1]

Game 2: Goodbyebluesky / taken over by et (Axis) vs gasteris (Soviets) ---- forum AAR: https://www.matrixgames.com/forums/view ... 2&t=402253

In this game, by January 1944, there were 42 Axis leader deaths (1 executed). The game is not close to over btw in this case, so many more leaders will die before it is done.

For this one, I have a complete screenshot in 2 bits of those leader deaths:

Although the overall number of dead leaders is about 20 (!!!) lower, it is still obviously similarly absurd:

L2.png (17.49 KiB) Viewed 1403 times

L3.png (756.73 KiB) Viewed 1403 times

L4.png (60.57 KiB) Viewed 1403 times

Some of the names that stand out there:

Nehring
Weiss
Fretter-Pico
Herrlein
Kleeman
von Kluge
von Knobelsdorff
Rommel
von Vietinghoff
Graesser
Hollidt
Guderian
Hube
Raus
Model
Rinhardt
Kempf
Halder

Sheer madness!



In addition, here are ones that died by turn 82, when et took over the game from goodbyebluesky, quote taken from discord AAR discussion:

et: these are the leader deaths that happened before under blues administration by turn 82:

L1.png (342.51 KiB) Viewed 1403 times

et: i lost halder to an execution
et: and like a few 7s like hollidt
et: @Goodbyebluesky were these mostly heart attacks or displacements
Goodbyebluesky: von Kluge was displacement
Goodbyebluesky: I remember him
Goodbyebluesky: I did check but I think thats mostly heart attacks. I was pretty dilligent on HQ placement
Goodbyebluesky: man some really good ones in there
Goodbyebluesky: yikes

[Beethoven1]

You may be wondering how this sort of thing happens, and why this is so different from what presumably happens in single player games or AI vs AI games. I think the reason is probably that a human player is much better at doing a large scale coordinated offensive with deep thrusts (which displace lots of HQs) than the Soviet AI. These 2 screenshots are from the David vs Caedus game:

leader1.png (129.92 KiB) Viewed 1391 times

leader2.png (3.55 MiB) Viewed 1391 times

In addition, human players will often use suicide cavalry or suicide 50 MP 1000 man tank brigades to suicide deep behind the AIs lines, where they will inevitably displace HQs. Here is an example from the gasteris-et game:

snake.png (1.28 MiB) Viewed 1391 times

And here is a more broader-front push example from the gasteris-et game:

snake2.png (1.45 MiB) Viewed 1391 times

Ideally, if a city such as Smolensk or Kursk is 10 hexes behind the line, the Axis player ought to be able to put army HQs there and they ought to be safe. But in the game, Soviets can often just snake high MP sacrificial suicide units far behind the lines to those places, not only cutting rails, destroying depots, and overrunning airfields to a ludicrous degree, but also displacing HQs and assassinating high ranking leaders along the way.

In games with the AI, I don't think anything like this happens, but this sort of nonsense is a pretty routine occurrence in multiplayer games (although it does vary from game to game depending on the tactics of both players).

Historically, of course there were men and reserves in rear cities like Smolensk. But these do not exist in the game, in part because the Soviet garrison TB abstracts those men so hey simply do not exist as counters on the map, and also in part due to the fact that 1 week long igo-ugo turns artificially delay the response that quickly organized Axis kampfgruppe reserves would make to Soviet attempts at deep attacks.



Given this sort of Soviet mobility and the lack of enough Axis counters on the map to provide any sort of depth against this sort of ridiculous snaking, the Axis is left with nowhere to put their HQs.

Axis can put their HQs in the rear, but they still get displaced.

Alternatively, Axis can put their HQs closer to the front but stacked on top of units. Sometimes this can save them, but many other times the Soviets can simply choose to attack the unit with the HQ stacked on it. Attack it with a 100-200k man attack, and that will also displace the HQ.

[Beethoven1]

Meanwhile the situation is often (but not always, depending again on player skill and tactics) totally different on the Soviet side.

Here are Soviet leader deaths on turn 102 of the David-Caedus game:

sovietdeaths.png (302.96 KiB) Viewed 1386 times

You can see these are obviously way lower, and are mostly made up of the deaths that are baked in on turn 1-3 or so.

They are also all pretty much all bad leaders.

The reason for this difference is pretty simple. Soviets have many more counters to work with than Axis, so they can have more units behind their lines protecting HQs. Axis can also obviously not afford to suicide units in the way that Soviets can suicide a 2000 man cavalry division or a 1000 man 50 MP tank brigade.

Axis also in general cannot do deep wide scale attacks, thrusts, or substantially large pockets, at least after the first few turns. In some rare cases this can be possible, but it is only really doable for very few Axis players, and basically you have to have a very good understanding of what Soviets are capable of and where their units are to be able to pull anything like that off as Axis.

Meanwhile, even a beginner can do that by the late game as Soviets. And moreover, Soviets can do that even with a very small OOB as compared to normal/historical (as in the gasteris-et game, which is currently in 1944, with barely more than 4.3 million Soviets on the map and 4.9 million Soviets in the entire Red Army). The small OOB makes things harder for Soviets, butt certainly does not stop them from being able to conduct a mass assassination campaign.



In most multiplayer games, it is almost as though the Soviets have a bunch of these flying around in the Axis rear constantly hunting generals and HQs in 1941-43:

[Beethoven1]

All of the above has been mostly about leader deaths from HQ displacement. However, in addition to that, the purely RNG deaths, or "heart attacks" that occur in the logistics phase are also too high.



As one example, in a game I have been playing as Axis, this happened as early as on turn 5, one of the relatively good German army leaders died for no apparent reason:

busch.png (22.97 KiB) Viewed 1376 times

He was about 10 hexes behind the lines, and there were no Soviet units anywhere near (even in Soviet territory, the immediate area was abandoned). The VVS also conducted no bombing missions - certainly not in the area, and not even anywhere else on the map, so it is not like Busch got hit by a stray bomb. Another rationalization for these leader deaths that sometimes gets made is "partisans," but this was only turn 5, and he was in the Baltics, where the Germans were initially greeted as liberators. Since it was only turn 5, there was not even any Soviet garrison TB yet.

Fortunately, we were playing PBEM with a house rule that in these sorts of cases of stupid ahistorical leader deaths, we re-run the logistics phase until there are no "heart attack" leader deaths. So my opponent graciously just re-ran the end of the turn and sent it back to me again.

However, you of course cannot do that in server games. Ability to override random leader deaths is currently probably the top reason to prefer PBEM over server games.

In this case, I could have gone on with the dead leader without being too unhappy (and the replacement was actually an upgrade), but the same thing can easily happen (way too often) to leaders like Manstein or Guderian or Zhukov or Vasilevsky. And that is just game ruining from a fun perspective. It ought not o be a thing, at the very least for the top leaders that did not historically die. Just so players can have fun playing with their leaders, which is a big part of the role-playing aspect of the game.



Here are some random quotes from discord AARs, showing that most players think these heart attacks are excessive/ridiculous, especially when applied to top leaders:

stephan: and in my soviet game temoshanko or whatever his name is had a heart attack

seristal: Speaking of LG and leaders, Purkaev suffered a heart attack in the late summer which was dismaying. I don't recall any Axis leaders biting it yet.

jango: inb4 you lose von Manstein to a heart attack 2 turns from now.

et: more importantly im basically out of 7 mech rating generals due to heart attacks and i think 2 displacements

albertN: Also is it known if the formula that kills leaders has been accrued? By logistic phase of turn 6 (so after 5th Soviet turn) I've lost 4 leaders as German, without HQ displacements.

ringloth: poor man died of a heart attack from stalin telling him he needs to relocate his hq after moving around hoovering up rifle brigades

Beethoven: Stamb sent back the turn. There was a heart attack (Lemelsen).

Zebtucker: Wow guderian had a heart attack

Jango: Fucking heart attacks. (von Salmuth)

neaugustus: Hube and Reinhardt didn't make through this mud season. both died to heart attacks

Veterin: in our game? can't recall if he's dead or not. i've lost a lot of good leaders. half of them from heart attacks (Manstein)

deaniks: I've had a lot of heart attacks this game (Rokossovsky)

Veterin: first random leader death on a good leader (Hube)

todger: unfortunately guderian died of a heart attack

neaugustus: second game in a row Tolbukhin dies because of heart attack

fulmen: Ya'll should just play with a mod that decreases the chance of leaders dying randomly
Jango: Can't mod hardcoded things.

[CaedusZ]

Very good post. There is little left to add for me as it covers all the experiences I made too.

In my game vs. Stephan Guderian got a heart attack as of turn 2! I'm glad, this game was PBEM so we were able to redo the logistics phase as for a server game that would be a serious issue for the whole german campaign.
Given the importance of leaders losing the few good ones really hurts. Especially when you are opposing a skilled player, who actively goes for HQ snipes, as it's pretty easy to guess where an HQ is stationed and which leader to expect there.
This also induces the current meta of soviets abandoning the south, as german breakthroughs are very likely and you're risking losing lots of good leaders to the german mech forces. Take this and heart attacks, and you've got a pretty ahistoric situations in the chains of command by game's design.

In my opinion there should be an option to turn off the random leader deaths. Also for hq displacements the current chance of 15 % stated in the manual feels more like it's actually 30% for a single hq.
Also keep in mind, in many late game turns soviets can easily displace five german hqs (or even more, but let's keep it simple). So statisticly speaking 0.85^5 leads to 55 % probability of at least one leader death this turn.

[Beethoven1]

Here are some historical statistics for comparison on Axis leader deaths:

https://forum.axishistory.com/viewtopic.php?t=135564



Most general deaths were division level commanders. Those are not in the game. So there are only 26 of generals who commanded corps or higher. Moreover, that is for the entire war and for all fronts, not just for the eastern front 41-45.




Causes of deaths seem to be tilted towards:

1) Front line combat deaths. These should only really happen to generals who are on/visiting the front.

2) Air attacks - however, keep in mind this is for all fronts and for all years. I would think a large share of those air attack deaths would have occurred on the western front in the late war, given the overwhelming Allied air superiority in he late war (for example, rommel's near death air attack). You would think that this should not be happening that often on the eastern front at least in 1941-42 and to some extent 43 when the Axis had air superiority and the VVS was in a shambles.


There are very few deaths by partisans there, only 5. So not that many deaths of generals far behind the lines can realistically be rationalized as "partisan attack."

[Beethoven1]

gasteris shared his Soviet leader deaths as of early 1944 from the game vs et (earlier goodbyebluesky). There are 39 deaths:

gasterissovietdeaths.png (905.25 KiB) Viewed 1271 times

This is somewhat the exception, normally Soviet leader deaths will likely be lower than this.

The reason for the high deaths is that this was the relatively rare game where Germany did actually get a more historical style number of pockets in 1941, and in addition gasteris was pretty haphazard with HQ replacement and not really thinking much about leader deaths I don't think.

So take this as probably about the upper bound on what Soviet leader deaths can get to in a player vs player game.

[Joel Billings]

Appreciate the feedback. We are looking at adding some new game options, and there's a good chance we can make a change to the formula and/or add an option. In recent AI test games, the Germans lose around 20 leaders through 1943, and 65 for the war. Soviets are around 20 in 1941 and 45 for the war. The AI isn't great with HQ placement, but it's also not targeting HQs. AI test games and human games can be quite different, and I follow your logic re the large number of General losses are division commanders, which is not what we're dealing with. Looking at the link, it would seem the number that is 26 German Army/Corps leaders killed, is a number that are killed in action. Another chart on the website adds in other kinds of deaths and has a much higher total (am I missing something), and remember, we don't deal with aging out of some of the older leaders so they have to die to be gone. Do you have a sense of how many of your leader deaths are due to the random factor, versus being displaced? I'm hearing that the random factor is very annoying, and you'd like to be able to turn it off (I think it's possible), but I wonder if the shear volume of leader deaths you see in 2 player games is due to displacements, which would require a different kind of change and/or option. I think we can do something here, so the more info you can bring the better we can try to deal with it. Thanks.

[M60A3TTS]

My question would be if a HQ is initially displaced and is co-located with one or more routed units (and no un-routed units), is that HQ considered to be alone in the hex, and therefore carries the higher leader loss chance percentage? That would accelerate the losses as the front lines open up and exploiting units maneuver through the gaps. It may be simplest just to have the lower chance of leader death in the event of displacement in all circumstances.

[Joel Billings]

Thanks for the question. I'm not seeing any rule about being alone in the hex causing an increased chance of death. I see an increase for being isolated. Do you have reason to believe that being alone causes higher losses? I can ask about that if you think this could be happening. Routed/depleted units are not considered combat units, so they would be of no value (just like if 2 HQs were stacked). Here's what I see the percentages are for displacements:

Errata – 01.02.21
The baseline chance a displaced/relocated HQ will lose a leader was changed, from 15% (50% if isolated), as follows:
• June 1941 – 5% (10% if isolated)
• July 1941 – 10% (15% if isolated)
• All other dates – 15% (30% if isolated)

Clearly the Killed in Action text is misleading as to what some of these non-displacement/non air-strike deaths represent.

[M60A3TTS]

Thanks for the question. I'm not seeing any rule about being alone in the hex causing an increased chance of death. I see an increase for being isolated. Do you have reason to believe that being alone causes higher losses? I can ask about that if you think this could be happening. Routed/depleted units are not considered combat units, so they would be of no value (just like if 2 HQs were stacked). Here's what I see the percentages are for displacements:

Errata – 01.02.21
The baseline chance a displaced/relocated HQ will lose a leader was changed, from 15% (50% if isolated), as follows:
• June 1941 – 5% (10% if isolated)
• July 1941 – 10% (15% if isolated)
• All other dates – 15% (30% if isolated)

Clearly the Killed in Action text is misleading as to what some of these non-displacement/non air-strike deaths represent.

Maybe I just imagined that, unless it was a WiTE1 thing. Or maybe it was the isolated condition I thought about and got confused.

Anyways, I just conducted a test where on turn 2 of the GC, 10 non-isolated Soviet corps HQs were placed on top of a combat unit and adjacent to them were 10 army HQs where they were by themselves. I then ran several German units into the Soviet combat units, displacing all 20 HQs. Only one army commander died. That's 5% which matches up with your numbers. If the 15% for all other dates is true, then by that math in the screenshot of David vs. Caedus with 8 German dead commanders, over 50 HQs were displaced. Are a lot of these HQs 1-2 hexes from the front lines???

Dead Soviet Generals in my game vs. jubjub, turn 41.

[CaedusZ]

I'm not sure how the game calculates the odds for tripple stacked HQs, I often felt like I would kill all three leaders there.

As it was a server game, I cannot access the older turns and my memory isn't the best. There certainly were some issues with HQ placements, but most german leaders died after the war turned in december 41 as I was able to break through his lines repeatedly. I got 62 kills by turn 102. Pretty sure, it was more than 70 by turn 120 as the war ended. Also not sure if these are only the german ones or wheter romanian, italian and hungarians are included. If so then they represent only a tiny minority as I conducted most of my offensives in the north and center.

I mean look at Beethovens 3rd post here or some further screenshots by me. (I'm pretty sure I did not displace 50 HQs, maybe 20 at max)

Leader Kills T32.PNG (49.66 KiB) Viewed 1160 times

Leader Kills T90.PNG (520.56 KiB) Viewed 1160 times

That being said my game is special as it was onesided from very early on. So the Gasteris game should be a better example, but even there by January 1944 42 leaders died, even though they were facing a relatively small red army.

From my experience the vast majority of kills comes from displacements, but heart attacks are the more annoying ones as you can at least blame your own actions of the last turn for displacement kills.

[Wiedrock]

I'm pretty sure I did not displace 50 HQs, maybe 20 at max

15% of 20 means you'd assume to lose/kill 3. The list shows 4. So if your guessed 20 displacements are correct, the 15% the Devs chose as an chance I could imagine to see in this example.

[CaedusZ]

I'm pretty sure I did not displace 50 HQs, maybe 20 at max

15% of 20 means you'd assume to lose/kill 3. The list shows 4. So if your guessed 20 displacements are correct, the 15% the Devs chose as an chance I could imagine to see in this example.

The 20/50 HQ discussion referred to Beethovens third post here, were I got 8 Leaders. The other screenshots were from different turns in which I at least according to my memory also displaced less than 20 HQs where the 15 % would make sense. But I think those leaders died in battles with the HQs really close to the frontline. Not sure if there is a additional chance of them dying in battles before the hq is displaced?

[Mehring]

It's not just deaths, though the game records them as such. Of the 151 at start german leaders available, just over a third were no longer available by the end of the war for various reasons. https://www.matrixgames.com/forums/view ... 3#p3687313

[Wiedrock]

It's not just deaths, though the game records them as such. Of the 151 at start german leaders available, just over a third were no longer available by the end of the war for various reasons. https://www.matrixgames.com/forums/view ... 3#p3687313

In theory there's a "end date" for many Leaders set in the Editor (the historical death/capture/arrest dates). Not exactly sure how that is being processed and when/how it is actually applied (I could imagine stuff like increased chance to die randomly after this date or so).

---

Finally, there is a small chance that leaders may be killed due to other enemy action.
The probability of this occurring is related to the distance the leader’s headquarters
unit is located from enemy units, with headquarters units closer to enemy units having
an increased chance of having their leader killed. Any leader in an HQ that is more
than 10 hexes from the enemy will have his chance of being killed reduced by two
thirds.

maybe some of all this "surplus" of deaths is related to this rule. Since if I afk a game I get this in one test-run till turn 10 (one Rumanian):

leaders_afk-test.png (75.28 KiB) Viewed 993 times

[Joel Billings]

AFAIK, the leader end dates are only there to prevent leaders from being loaded into scenarios that start after the leader was no longer available. Don't think anything once the scenario is loaded is using the dates to alter the chance of death. The distance to the enemy is a big factor in what we call Leader Attrition, which is loss of leaders not directly tied to displacement of HQs.