Victory conditons: possibilities + the burden of coding a CPU opponent

[Lt Bull]

I was originally going to reply to Skolman's post " new victory conditions ", but decided to start a new thread about it.

If there has been on aspect of CM that has not changed in the last 6yrs since CMBO, it has been the abstract concept of victory conditions and how they are defined/determined.

At the end of a CM battle a "score" for each player is generated base on the points from:

"casulaty points"- based on fixed scores for each type of enemy unit destroyed/captured

"flags held"-based on who is deemed to control each of the "flags" on the map at the end of game

"exit bonus"-based on a fixed score given for each specified unit exited off teh map

"scenario bonus"-a scenario designer specified score bonus given to one side or both (I think).

The scores of each player are proportionally compared and the differential assigned an outcome grade (draw, minor, tactical or major).

As we know, the parameters for "victory" in any real battle are not always the same or graded/measured in the same way. The defintion of victory can vary quite significantly depending on the tactical/strategic situation at hand. Ultimately, it is based on the orders/expectations of the opposing commanders or even vary by the dynamics and progress of the battle.

A few suggestions on how the current "score" concept could be enhanced just to get people thinking (note: these all could be scenario design options that could be switched on/off by the scenario designer/QB settings as required):

# At the unit selection stage you might allow players the option to spend MORE on units than the noiminal "purchase points" alloacted to them by the scenario. In doing so they incur a "victory score penalty" proportional to the ammount of "purchase points" spent in excess.

# On the other hand, the player also has the option to spend LESS on units than the nominal "purchase points" alloacted to them by the scenario. This would give that player a "victory score bonus" proportional to the "purchase points" not spent.

This could simulate a local commander requestng more/less troops from the pool of units available knowing that the level of victory can sometimes be measured against the number of units committed to an action to achieve some objective.

This could add a new dimension to the game where players can gamble up or down before the battle to spend more or less on units, depending on how they size up their opponents and the task at hand and give the feeling of being involved in a "bigger picture" battle.

# Allow the option to have flags (or victory locations (VLs)) to generate "victory points" for the player for each turn they control them. This may go some way to eliminating the mad "last turn VL scramble" you can see happeneing in the current game. This can simulate the tactical importance of holding ground based over a period of time rather than a a precise finite point in a battle (ie. the end).

# Allow the "points per turn controlled" for these flags to vary depending on what turn it is (or even some other factor!). The possibilities and implications for this could really vary and direct the dynamics and flow of a battle.

eg. this could work really well to add pressure to the attacking player moving forward trying to secure VLs before their "points per turn" value drops to even zero while the defending player decides when to abandon a forward defensive VL position as it's "points per turn" value gradually diminishes each turn. This could simulate the expectations that an attacking commander might have of the expected rate of advance or a defending commander the expected rate of fallback withdrawl.

# Allow the VL points value to be different for each player.

eg. a VL might be worth 100 points if player A controls it or 300 points if player B controls it. This could simulate the differing tactical significance of partiular VLs might have as seen by the opposing command HQs.

# Alternatively, allow the option to have the "points per turn" value of VLs to start changing based on "triggers" such encroachement of particular enemy units within a certain range of the VL. This could simulate being rewarded for holding a tactical objective made more valuable by the mere fact the enemy have decided to try and capture it ie. encroached within a certain range.

# Allow the option to have points awarded for VL ownership to be proportional to the number of friendly units occupying/controlling the VL footprint. This could get away from current gmae concept that a VL controlled by one sniper, for victory considerations, is no different as far as scoring goes as the same VL being controlled/occupied by a whole platoon/company etc. This could simulate victory based on a level of control.

# Allow the option for VLs to have "footprints" of various sizes.

# Allow the scenario designer the option to assign "bonus kill points" for destruction of particular units. This could simulate a tactical situation where loss of particular units would have severe implications strategically despite the local/immediate tactical worth of the unit.

# Allow players the option to "request for additional reinforcements" while incurring a "victory point penalty" depending on when they request them. Perhaps a random probability of the reinforcements ever arriving could be used. Perhaps the "victory point penalty" is applied as soon as the request has been made (regardless of whether the units actually arrive) or only IF the units arrive, or some combination of the two.

# Allow the "map exit point" values for partiuclar units to vary depending on what turn they exit. This could simulate answering a request by HQ to send units to a reinforcement pool.

# As a totally alternate optional victory system, use a varaiant of the very clever and dynamic victory determination system used in Battlefield:1942 where each side starts off with a certain amount of "victory points" to "defend". Victory points are deducted from the opposing players current total immediately for the destruction of an enemy unit and for the duration of time you occupy a VL. When the turn limit is reached, the side with the most victory points is awarded a level of victory based on the differential. If one side loses all of their "victory points" before the turn limit is reached, victory is immediately assigned to the player who still has some victory points remaining.

The standard "score" system as a whole works well, but it does have it's limitations when you consider the range of other "real life" victory parameters that scenario designers may want to include.

The abstract concept defining victory in CM using the "score" system could be developed further. It is also possible to consider a system which gives the scenario designer the option to use a composite type of system for determining victory conditions based on very precise and scenario specific "tactical conditions". In these kinds of battles, the "tactical conditions" are either fully accomplished or not accomplished

eg. Occupy Hill 203 with at least one Inf Co and two tanks for "x" consecutive minutes while preventing any enemy armored incursions within "x" many meteres of the objective.

On top of this you can "layer" additional "tactical conditions" to spice things up.

eg.

- destruction of all enemy armor will result in immediate victory

- greater than 30% infantry losses/20% armored loses will result in immediate defeat

- destruction of HQ unit will result in immediate defeat

- enemy exiting 3 or more tanks of map edge will result in immediate defeat.

The possibilities of what the "tactical conditions" are well defined. In fact, there is nothing about the game right now that could prevent players (barring FOW) from "manually" keeping track of these very scenario specific/finite "tactical conditions" to determine victory. Would just be good if the game could track these very simple conditions for you.

As determining victory conditions in any game is really just a passive "book-keeping"/number crunching exercise that gets performed once at the end of every turn, I wouldn't think it would be too hard to mechanically incorporate some of the victory conditions to be tracked/monitored in a game like CMx2 which is being built from the ground up.

However, this may not be why BF may/will not change the current "victory determination system" as it is in CM going into CMx2.

I think the aspect of modifying the way victory conditions are determined/achieved in CM (or any other suggested game feature) is HUGELY limited and influenced by BF having to code a CPU opponent that can interpret, understand and react to achieving these "complex" victory conditons (complex for a CPU to interpret at least).

The effort to implement cool game features that can be handled by humans might be a sinch and HUGELY appreciated by human players, but to then have to program the CPU to be good at playing along to these new rules..... :eek: !!!

The burden of coding a competent CPU opponent to control/direct a digital army to achieve a set of victory conditions easily understood, interpreted and appreciated by humans. Is it holding back game design? I will discuss further on next post.

Lt Bull

[ August 27, 2005, 07:20 AM: Message edited by: Lt Bull ]

[Lt Bull]

see this post.

[ August 29, 2005, 07:43 AM: Message edited by: Lt Bull ]