Illegal charge

[tyronec]

An odd occurrence in a current DL game.

1. The Thureos charge the Peltasts who evade.
2. The Peltasts get a long way away.
3. The Legion charges the Peltasts in the rear, form 22 MPs away !

I have a video if someone wants to look at it, just PM me an email.

c.jpg (367.4 KiB) Viewed 1986 times

[SimonLancaster]

Did the option come up to charge the Peltasts with the Legionaries and you pressed charge or did it happen automatically?

[tyronec]

You would have to ask nyczar, he is playing the Romans.

[nyczar]

That is weird, that not what I saw at all. The skirmishers in my view evaded to where the "18" was in my version. Easily in reach of my legion.

[tyronec]

Even more strange. For that to happen they must have done a second evade move, first from the Peltasts and then from the Legion.
But in the playback video they did not go there.

[rbodleyscott]

That is weird, that not what I saw at all. The skirmishers in my view evaded to where the "18" was in my version. Easily in reach of my legion.

Even more strange. For that to happen they must have done a second evade move, first from the Peltasts and then from the Legion.
But in the playback video they did not go there.

The Replay is not a video, it works by setting the game to its status before the other player's turn, then runs through all the moves that occurred. This reconstructs the situation at the end of the other player's turn. The situation at the start of your turn is actually the situation created by the Replay.

If for some reason it gets out of synch, the "playback" won't match the actual events. In this case the peltasts presumably ended up in a different place in the original turn, within charge reach of the legion, but then in fact moved further in the replay. The replay then activated the legion charge, ignoring the fact that the peltasts are now out of reach by the normal rules of the game.

The commonest reason for the replay to go out of synch is if the players' script builds differ in any way. The replay cannot work correctly if this is so, and it isn't a bug if it doesn't work correctly when the scripts differ. MP players should not mod their builds in any way.

It is, however, possible that the script builds do match, but the replay went out of synch for some other reason. If so, that would be a bug, but it would be one that is very hard to track down.

[tyronec]

I still cannot make sense of this.
In the video (I understand that the playback is not a video, am referring to my video of the playback) the Peltasts are charged by the Thureos and then by the Legion. Two charges are impossible because either the Peltasts are within range of the Legion and should not evade again OR they are out of range and cannot be charged.
The only legal way I can see for the situation to be created would be if the initial charge was by the Legion. That way maybe both units moved 4 squares. But that would mean the Thureos don't charge while the video shows that the do charge.

Don't know what you mean about modding builds, can players edit the game in some way and still play MP ?

Anyway, it is not having much impact on the game so I will just play on.

[rbodleyscott]

The only legal way I can see for the situation to be created would be if the initial charge was by the Legion. That way maybe both units moved 4 squares. But that would mean the Thureos don't charge while the video shows that the do charge.

The point is that it doesn't have to be legal for the replay to activate the charge; the replay doesn't go through the game logic to see if the charge is legal, it just activates the charge on the target unit that was recorded in the list of actions for the previous turn. If everything is working as it should, this produces exactly the same result as it did in the other player's turn. But because the target unit in this case isn't where it should be, the chargers move an excessive distance to charge it in its new position.

Don't know what you mean about modding builds, can players edit the game in some way and still play MP ?

They can, but they should not, because it messes up the replay. The game will issue a message to the other player if the scripts used for the previous turn are different from the scripts on his machine.

[nyczar]

not that it matters now, but it was the Thureos that charged first. then the legion when I saw I could catch the peltasts.

[pipfromslitherine]

Note that this can (and likely often does) happen when versions update. If the script logic is slightly different between the turn completion and the replay then the outcome may be different. The vast majority of the time the difference is likely not to be noticeable, but it could be more egregious.

Cheers

Pip

[tyronec]

From what nyczar has reported I think something is not right here.
As the Thureos charged first the peltasts must have been within range of the Legion. Since they have already evaded once they cannot evade again and should have been caught nearer to their original position.
It seems both Legion and Peltasts have been teleported somewhat.

However is doesn't really matter in terms of the game and I am merely reporting this as a point of interest.

[rbodleyscott]

From what nyczar has reported I think something is not right here.

Nobody is denying that. We are just trying to explain how it happened.

As the Thureos charged first the peltasts must have been within range of the Legion. Since they have already evaded once they cannot evade again

I don't think they did evade twice. If they threw up on their VMD their move would be 16 AP. The first diagonal move of a flee move is treated as costing only 4 AP (to avoid issues with units being more likely to be caught on diagonals), hence a diagonal flight move of 3 squares is treated as 16 AP.

So the only actual anomaly is that the flee move of the peltasts was different in the replay from the original turn.

This will almost certainly have been due to the random roll for the VMD being out of synch, which could either occur if the scripts were not identical, or due to some as yet unknown bug.

The excess move distance of the legion is due to the replay simply activating the recorded charge move against the peltasts, without any of the checks for permissable distance etc. that would be carried out in an actual turn. This would work correctly if the peltasts were in their "correct" position, but because they have moved further in the replay than they did in the actual turn, the legion charge move is excessively long.

[tyronec]

There is no problem with the Peltasts getting to their final position from evading the Thureos, that is fine.
However if they did evade to there then they are out of range of the Legion, so the Legion could not have charged them.
If they were in range of the Legion, as nyczar reported, then the final position should be different to what it was at the start of my move.
Both the video and nyczar report that they were indeed charged by the Thureos. Either both nyczar is mistaken and the Charge shown in the video did not happen OR both units got teleported
somehow.
Maybe I am mistaken in this and am misunderstanding something...

It has probably all worked to my advantage as when the Legion later routed the Peltasts they were out of position from the front line.

[rbodleyscott]

There is no problem with the Peltasts getting to their final position from evading the Thureos, that is fine.
However if they did evade to there then they are out of range of the Legion, so the Legion could not have charged them.
If they were in range of the Legion, as nyczar reported, then the final position should be different to what it was at the start of my move.
Both the video and nyczar report that they were indeed charged by the Thureos. Either both nyczar is mistaken and the Charge shown in the video did not happen OR both units got teleported
somehow.
Maybe I am ...misunderstanding something...

I fear so. I will have another go at explaining it.

In Nyczar's actual turn:

1) The Thureophoroi charged the Peltasts.
2) The peltasts threw down on their VMD and evaded 1 square. (Ending in the square with the 18 casualties displayed on your first screenshot).
3) The legionaries charged the peltasts who were in charge range.

Nyczar then ends his turn which is then saved to the server. What is actually saved is not the position at the end of his turn, but the position at the start of his turn, plus a list of game actions that occurred during his turn.

In the replay:

The replay starts from the position at the start of Nyczar's turn, then runs through all the recorded actions without testing whether any of them would normally be permitted. If everything is in synch, then this will reproduce the events of Nyczar's turn exactly, so there is no need to test the legality of actions as that was already tested in Nyczar's turn. The situation re-created by the replay is the situation that will apply at the start of Tyronec's turn.

So in the replay:
1) The Thureophoroi charged the peltasts.
2) Somehow the RNG has got out of synch, the peltasts throw up on their VMD instead of down, and evade 3 squares instead of 1.
3) The replay activated the legion charge on the peltasts that was recorded in the saved data, without checking whether the charge is still legal, and the legion scoots 22AP after the peltasts.

[tyronec]

Ah, now I understand !
Thanks for your patience and detailed explanation.

[nyczar]

I would like to buy some teleportation ability for future matches....42 force points seems right.

[SimonLancaster]

Did either of you edit the game in some way or use a mod?

[tyronec]

Not me.

[nyczar]

No.