Overwatch, Attacks of Opportunity

[ushas]

All me is on the reaction radius side here. 

Yeah, also thought about low CTH shots at distant enemies, now I see what you meant with the minimum CTH, but if you use this for permitting reactions it's harder to guess for players (also some will whine that it's too low and some too high). So we may end up using only highly skilled ones, more so if can't control at who they shoot.

Whereas in case of something more simple like discussed distance being the threshold you have more clear base to build upon your positioning and to factor it in your decision whether to postpone your shot into overwatch or not.

Let me know if you're not diverted enough from the CTH threshold, have reserved one more argument as reaction fire.

Distance is usually based on some stat, but I like the spirit of presets idea. Giving some more control. Just not sure about usability vs. implementation yet (perhaps offer some slides in options, but that's also more work). If I may, depends what you want to accomplish - for similar principle it's also possible to use the weapon+mode combination itself as the distance threshold defying (aka loading a constant variable from an array). Pistol in snap reacting at lower distance than in normal mode, precision rifle having all distances bigger than pistol, a shotgun has it already defined, etc. This one loosely gives AP cost vs. distance correlation and is weapon dependent.

Ok, overwatch button makes a sense. Kudos for reasoning, also tells us more about your intentions. Just don't forget a key-binding. Btw. reaction radius can also be shown as highlighted on the ground when mouse-overing that button.

Rest later, am much slower to type.

[Daithi]

I'm leaning toward CTH being the triggering factor, with 4 settings: 0-25%, 26-50%, 50-75%, 76-100%. It's likely that players will quickly get a feel for a good range, appropriate to the weapon and skill combination. Distance modified by fire mode is from a coding perspective more complex, and am not convinced a range for trigger is any better than CTH for trigger: it mostly swaps one unknown for another (unknown (when overwatch triggers) -> unknown (chance of attack being successful).

By all means, deploy your argument Won't get to it today, and would rather get ideas on it before committing.

[ushas]

By all means, deploy your argument

With pleasure
Unfortunately, my turns are delayed so it will be in the next one. Keep the shields up!


Might be wrong, but I believe who and when fifes does impacts the decision making.
So after vanishing the limitation is it now like everybody at their first trigger but sequentially? (what happens if one of them eliminates the target?)

I don't think it will have much impact on how the system works. Currently set to melee first, then ranged, in order that characters entered overwatch.

Perhaps may work with delaying actions or if being asked that y/n, otherwise other options would give better prospects, for example, the CTH order or distance from the target (both similar, CTH favors more real chance, while distance is again easier to use when you can guess the enemy's position)...

I suppose also boils down to additional calculations and programming though.

Feeling is that overwatch should be automatic, with some control over when to trigger. Otherwise, control will continually pass to player (on every tile the enemy moves), they will decide whether to take the shot or pass. The second approach gives more control, but it's going to undermine pacing and add a lot of clicks, which is going to be frustrating.

Depends if you operate under "reaction fire given on each step". There are some options, but what triggers a reaction - simply each step on a tile?

In old x-com there is difference between   move then shoot   and   shoot then move. The former puts you into worse position at the shooting time as you already spent some TUs, thus having lower reaction score. Another example is UFO:AI with their reaction fire contest, see ufopedia.

So similar AP comparison mechanic:
Let's reserve some AP amount for overwatch (can simply remember the mode cost). Next it's an enemy's turn.
-- Till his AP spent is lower than our reserved -> he acts still faster.
-- Once his AP spent is higher than our AP reserved -> reaction fire!
Note1: Obviously, makes snap earlier to proc, but if an enemy is coming to us - using normal mode will proc when he is more closer (-> differs in CTH).
Note2: No need to evaluate costly conditions like cth or distance until the acting unit spends more AP than is the lowest AP reserved on the opposite side. 

Back to the question when a reaction procs. Triggering closer to the beginning of enemy's turn may look good, but in case already inside the threshold (be it CTH or distance) that also pose the question why I wouldn't rather fire during the proper turn before as the conditions aren't so much different (eg. he steps just one tile, not necessarily in my direction)...

Obviously you've thought such situations over, even thinking to give us an y/n option. However, couldn't be choosing the weapon+mode and threshold on its own enough for declaring how we wish to trigger? The AP comparison above and sequencing watchers by CTH are examples how to do that. Yeah, asking rather to be constrained than y/n on each step.     

PS: Not sure but based on what you said, the Overwatch button might work as the reservation. You can reserve weapon+mode & AP any time, so you don't accidentally spend AP or have the wrong weapon readied. Problem: it would require (un)locking the weapon on one of the hands when (un)reserved.

[Daithi]

Might be wrong, but I believe who and when fifes does impacts the decision making.
So after vanishing the limitation is it now like everybody at their first trigger but sequentially? (what happens if one of them eliminates the target?)

They will stop firing/attacking. The limitation was initially to prevent all characters firing at once and wasting attacks on a KO enemy.

Perhaps may work with delaying actions or if being asked that y/n, otherwise other options would give better prospects, for example, the CTH order or distance from the target (both similar, CTH favors more real chance, while distance is again easier to use when you can guess the enemy's position)...

Not keen on the idea of asking the player at each step. Will really slow things down. Picking highest CTH first sounds good though. I'm probably going to move on and try to get AOO and initiative order working first.

I suppose also boils down to additional calculations and programming though.

Yeah, getting pressed for time. Strongly considered skipping this update and moving straight to prepping for steam.

There are some options, but what triggers a reaction - simply each step on a tile?

In old x-com there is difference between   move then shoot   and   shoot then move. The former puts you into worse position at the shooting time as you already spent some TUs, thus having lower reaction score. Another example is UFO:AI with their reaction fire contest, see ufopedia.

So similar AP comparison mechanic:
Let's reserve some AP amount for overwatch (can simply remember the mode cost). Next it's an enemy's turn.
-- Till his AP spent is lower than our reserved -> he acts still faster.
-- Once his AP spent is higher than our AP reserved -> reaction fire!
Note1: Obviously, makes snap earlier to proc, but if an enemy is coming to us - using normal mode will proc when he is more closer (-> differs in CTH).
Note2: No need to evaluate costly conditions like cth or distance until the acting unit spends more AP than is the lowest AP reserved on the opposite side. 

Back to the question when a reaction procs. Triggering closer to the beginning of enemy's turn may look good, but in case already inside the threshold (be it CTH or distance) that also pose the question why I wouldn't rather fire during the proper turn before as the conditions aren't so much different (eg. he steps just one tile, not necessarily in my direction)...

Obviously you've thought such situations over, even thinking to give us an y/n option. However, couldn't be choosing the weapon+mode and threshold on its own enough for declaring how we wish to trigger? The AP comparison above and sequencing watchers by CTH are examples how to do that. Yeah, asking rather to be constrained than y/n on each step.     

PS: Not sure but based on what you said, the Overwatch button might work as the reservation. You can reserve weapon+mode & AP any time, so you don't accidentally spend AP or have the wrong weapon readied. Problem: it would require (un)locking the weapon on one of the hands when (un)reserved.

Yeah, it is triggered when the character reaches a new tile. The AP trigger mechanism is interesting - wasn't aware of that system - it definitely adds more depth. You seem to have an encyclopedic knowledge of TB games! I'm going to run with the current setup for now, but will keep this in mind for tweaks. As you pointed out on dex, having LOS to most enemies and no FOW reduces the usefulness of overwatch, so will put the version out, and get feedback on whether it's a good area to work on some more.

[Nomad]

I'm leaning toward CTH being the triggering factor, with 4 settings: 0-25%, 26-50%, 50-75%, 76-100%.

Do you mean that if someone is set to 0-25% then they wouldn't attack someone if the chance was 88%?
Or do you mean that the settings would be 0+ (shoot at the first target in range that moves), 25+ (only shoot at targets that have a greater than 24% chance), 50+ (shoot if greater than 49 CTH) and 75+ (shoot at any target with a 75 or higher hit chance)?

[ushas]

Forgot to mention - the argument - it's in the burst mode.

Was thinking how both sides interact. So let's say the acting side knows or suspects the opposition is in overwatch.
Even then how well I actually can guesstimate where/when its triggering?

1) CTH: Not sure how easy is to know the current enemy's skill & weapon & threshold (they lvl up, the gang AI does upgrades etc) and compare with current positions and defenses of your allies, to evaluate their CTH. The same for the enemy side - unless they cheat, how they evaluate the threat? It does correlate with distance, but can change from target to target and between situations a lot. 

If I suspect to be under the reaction-procing, it pins down in acting (fully in case the first act always triggers, or partially for AP speed comparisons and such). As there is no defense - options are either eliminate all overwatchers or take a risk and move? Would be more paranoid with the blurriness of the CTH edge. What I'm sure is that moving away gives them lower CTH (what if more overwatchers around?). Would like also a guess whether triggers at all.

2) Distance: we see what weapon the enemy readied so have a fairly good idea on the reaction radius as this works for everybody the same (I assume Nomad's presents also +/- correlate with weapon types). The same with enemies, it will feel fair for them having this info now because we have it too.

Deeply inside the threshold there is no difference between using CTH or distance -> fairly sure to be pinned, but the closer the edge the more it differs what I know and thus can do about it... 

Overwatch doesn't need to be good only when fires and hits though. I use fire grenades more for their hazard feature - lower willingness of enemies to go in. You can even put the same invis. hazard on tiles under overwatch, perhaps with lower repellency. An enemy near the edge may decide to step outside and go around, so no reaction -> but then I've been successful in keeping him out. It has an effect without the trigger.

It's likely that players will quickly get a feel for a good range, appropriate to the weapon and skill

and to the target's evasion & armor and to the target's cover angle and to the player's crouching and on both sides to status debuffs & leader & perk modifiers combinations... I would run a script for that, a big fun, thanks!

it mostly swaps one unknown for another (unknown (when overwatch triggers) -> unknown (chance of attack being successful).

Isn't that whole swapping thing more near the edge of the threshold situation? If choosing +/- comparable thresholds, then yes in statistically significant repetitions of enemies near the edge we could see that difference. But in one case I would be less inclined to enter the overwatch.

Good fight! (your stance is sound, merely supplying the opposition hoping is helps somehow)

[Daithi]

Do you mean that if someone is set to 0-25% then they wouldn't attack someone if the chance was 88%?
Or do you mean that the settings would be 0+ (shoot at the first target in range that moves), 25+ (only shoot at targets that have a greater than 24% chance), 50+ (shoot if greater than 49 CTH) and 75+ (shoot at any target with a 75 or higher hit chance)?

Yeah, I didn't describe that very well, it's the latter. 0-25 will fire if there is LOS and the others will activate if the minimum threshold is met.

[Nomad]

If you're doing that, then you really only need the lower number (just an opinion).
0+, 25+, 50+, 75+

Edit:
What I mean by that is that otherwise it's:
0-100%, 25-100%, 50-100%, 75-100%

[Daithi]

@ Ushas: In the event there's enemy overwatch (unclear at the moment) it will probably trigger easily. The reason being that I think it will be difficult to make the AI use OW effectively, without the player being able to game the AI. If the AI just holds fire on overwatch, it will make it less impactful. Didn't mean to pull the rug out from under your feet with the enemy overwatch part of your reasoning - this was something I only started to realise last night 

Likewise, I think your reasoning is good, and appreciate you and Nomad taking the time to offer different perspectives.

@ Nomad: Yeah you're right, only minimum to trigger would be required.

[Nomad]

I would argue that the enemy has superior numbers and don't need Overwatch.
Or... I would argue that their use be determined by the selected difficulty level.  =)

[ushas]

Strongly considered skipping this update and moving straight to prepping for steam.

Will be there any other update before going EA?

The AP trigger mechanism is interesting - wasn't aware of that system - it definitely adds more depth.

Timing is powerful tool, fun to gamble with. Not the best, but comes to mind as simple to use, no other stat checks. Loosely relates to what ArchAngel mentioned on when to act.

As you pointed out on dex, having LOS to most enemies and no FOW reduces the usefulness of overwatch

By its definition - maybe, but can merely fulfill slightly different purpose here. As you said it has already some situational benefits. Apart that, in both AOO and Overwatch we undergo some risks, the former gives attacks for free, would it make sense for the latter if targets can't benefit from evasion? Or what if successful overwatch hit distracts the target (Distracted: -50% AP for 1 turn)? If works right away the unit's turn was practically disturbed.

In the event there's enemy overwatch (unclear at the moment) it will probably trigger easily. The reason being that I think it will be difficult to make the AI use OW effectively, without the player being able to game the AI. If the AI just holds fire on overwatch, it will make it less impactful.

Oh, but we always game it, resistance is gallant but futile.
I dunno. There is potential in side effects. Not all spells are damaging, some makes you unable to act, other run in fear.

Looking at large, the sides aren't equal, different strengths, different weaknesses. So still the same mechanic can include stuff impacting players more. Our units are more precious, so yeah easy to trigger helps, anything making the target's life miserable (eg. see above), etc.

We are very, like really very, susceptible to surprise. Even when easy to game, at least the first several times seems challenging just from the adaptation pov. You have variety in factions/ranks/traits going on, which is great.

On the AI in general I'm curious about tricks on the illusion of AI adaptation. Not sure how could go, probably just stupid, but perhaps some variables may vary in time, eg. that repellency of overwatch danger zones, etc. So at the beginning of the battle they fall for traps and are easy to game, lulling us into comfort, but later they gradually change behavior, favoring different approaches, more cautious, looking as if the consequence.

You can also observe and collect info on the player's play. Then use it against him within the battle or across the campaign. This can work as misleading too (not to be used often, of course). For example, a timely placed new bark. So even when the AI doesn't change at all, we would think it did somehow, even random outcomes attributing to that.

I would argue that their use be determined by the selected difficulty level.

Having the rate of selected tactics or even the use itself adjusted by difficulty sounds interesting!
Sigh, more work on balancing, isn't it?

Didn't mean to pull the rug out from under your feet with the enemy overwatch part of your reasoning - this was something I only started to realise last night 

Don't worry, one's strategy does not fall apart by circumstances. In fact, already recuperated forces. Stay tuned!

[Daithi]

@ Nomad : Am holding fire on enemy Overwatch for now. Difficult to call, but Implementation time might be better spent elsewhere. Will keep your difficulty suggestion in mind though

@ Ushas: At least 1 more update, maybe 2-3. They are likely to be shorter updates though. Need to rework char system, reduce enemies, add polish, stomp bugs. If V19 passes your and Nomad's scrutiny, it might be the first Steam version. It needs to be in good shape when it hits Steam though, so it is one of those "when it's ready" situations.

[Nomad]

Waiting before putting it on Steam is wise if you have that luxury.
You only get one first impression, so make a good one.

I think that once we get to the point where we can actually get info on the underbosses, bosses and facilities without it crashing, then it will be at a good place. If I'm not mistaken, past that point it's really just adding gear and features. The base game is pretty much there.

So, like you say, if the next couple of releases can clear up some of the bugs, then it should be looking pretty good.

[ushas]

Agreed. Knowing next to nothing about the steam, but based on what heard, having a well playable start could be an advantage. Later may afford break things a little bit here or there. I guess more backers will jump on. So some reviews.

Shorter updates seem reasonable, if it's not much hassle to put a build out. Those in-between turns features themselves sound as a big addition.

As said above, we haven't reached the 2nd part yet, so naturally worried about the boss-hunting too. Will do my best to help hammer the freezing bugs.

[Daithi]

Waiting before putting it on Steam is wise if you have that luxury.
You only get one first impression, so make a good one.

I think that once we get to the point where we can actually get info on the underbosses, bosses and facilities without it crashing, then it will be at a good place. If I'm not mistaken, past that point it's really just adding gear and features. The base game is pretty much there.

So, like you say, if the next couple of releases can clear up some of the bugs, then it should be looking pretty good.

Have no direct experience of Steam EA, but the developers I've spoken to gave me that advice, to hold of on EA until the game is in as good shape as possible. They also said for single player EA games, a long time in EA is counter productive. Releasing too late in the year isn't good either, very hard to get coverage, and could really do with getting Vigilantes on Steam, so there's multiple opposing factors at play.

Yeah, need to work on interrogations, so it's more stable, and to allow honing in on one lieutenant or facility. There are loads of little things that need to be tackled, for the most part these will be smaller tweaks. Will start compiling a list of these.

Agreed. Knowing next to nothing about the steam, but based on what heard, having a well playable start could be an advantage. Later may afford break things a little bit here or there. I guess more backers will jump on. So some reviews.

Shorter updates seem reasonable, if it's not much hassle to put a build out. Those in-between turns features themselves sound as a big addition.

As said above, we haven't reached the 2nd part yet, so naturally worried about the boss-hunting too. Will do my best to help hammer the freezing bugs.

From what I gather, bad EA launch will seriously hurt the game. Building itself, maybe 1/2 day, but several days testing are also needed. This isn't wasted time though, as it helps with stability. Yeah, changes like AOO, OW, Delay are high risk with bugs. If hadn't tackled them in this update, probably would been too cautious to add them. Depending on progress with V19, intel gathering could be tackled then. Freezing input bug is the biggest bug concern at the minute. Hopefully we can track that down soon.