There seems to be a great arithmetic deficiency in CTF players; hurling themselves at the enemy flag with reckless abandon.

Dave's Law of Flag Loss tells us that this is, in fact, a poor choice. Taking "f" to be the flag and "s" to be one's self, and considering enemy points to be negative values, we see that:

f loss = f pickup - f return - s death

This meaning that points received for a flag loss include the one point for picking up the flag, subtracting five points for the flag return, and subtracting another point for your death.

Therefore attempting to take the flag when one is not able to escape results in a net loss of five points. Doing this four times would be enough to negate the twenty points of a flag capture.

In conclusion, the next time you want to be a hero, assess your ability of escaping the enemy base with the flag. If you feel that you cannot do this, go read a book.

It's going to take a lot more than a forum post.

Pictures, videos and audio are needed for the masses.

I think the masses already know that a match can be won by base camping and successfully ambushing enemies right after they got ones flag.
This however could be a bit boring and one needs to rearrange the ambush often to void the attackers strategy changes. Additionally the whole team needs to be good at staying alive...

What happens if one dies or suicide in an area where the flag is returned automatically (space, lava, etc.)?

By the way if one always kills some enemies on ones way to the flag, problems disappear silently.

This is why I don't play CTF. Score system is simply wrong.
I will wait until 1 capture = 1 team point. And frags have nothing (very little) todo with the teams win or loss.

Yah, like in Q3. I don't like it when the campers can get the equivalent of a capture just by killing noobs.

Is it possible to set score mode with variables?

I never made a CTF server, but shouldn't that be configurable?

just set g_ctf_flagscore_capture to 1000, fraglimit to 10000 and it will be hard for campers to win...

edit: sry if i write shit, i'm fucking drunk...

Idiot me, I think I'm best at getting raped at only a few meters of the enemy base. Thanks for pointing this out .

Dave wrote:f loss = f pickup - f return - s death]

Do you have any experimental data to support this?

Dave wrote:Taking "f" to be the flag and "s" to be one's self, and considering enemy points to be negative values, we see that:
f loss = f pickup - f return - s death

Negative loss = positive?


R loss, t = total rate of loss per minute
R loss, 0 = standard rate of loss per minute
n fails/minute = average number of failed flag captures per minute
f pickup = points from enemy flag pickup
f return = points from enemy flag return
m death = points from team member death

Xeno's Corollary to Dave's Law of Flag Loss:

R loss, t = R loss, 0 + n fails/minute * [ f return + m death - f pickup ]

n fails/minute = average number of failed flag captures per minute

Xeno do you think that the number of failed flag captures pm is Gaussian distributed? Then the average would be representative, but, i think we should replace it with a random variable, and you should think a bit more and prepare the stochastic differential equation giving the more realistic formulation of your corollary. If we then integrate it with a nice Ito integral (Ito as the map from Red Ishad ) from 0 to timelimit we will get the total - score provided that we get reasonable statistical data for the volatilies of each players playing style.

phd training... worthless... i should remove the title chemeng phd for nexeng...

pain_fedora6 wrote:n fails/minute = average number of failed flag captures per minute

Xeno do you think that the number of failed flag captures pm is Gaussian distributed? Then the average would be representative, but, i think we should replace it with a random variable, and you should think a bit more and prepare the stochastic differential equation giving the more realistic formulation of your corollary. If we then integrate it with a nice Ito integral (Ito as the map from Red Ishad ) from 0 to timelimit we will get the total - score provided that we get reasonable statistical data for the volatilies of each players playing style.

phd training... worthless... i should remove the title chemeng phd for nexeng...

I had to do a Statistics class for Chemistry. Spent about 3 hours studying for it and passed. Totally worthless.

Evidently failed caps are abundant with... people that cant flag run, or the people that run to get the flag when they have 10 health left. Genious.

morfar wrote:This is why I don't play CTF. Score system is simply wrong.
I will wait until 1 capture = 1 team point. And frags have nothing (very little) todo with the teams win or loss.

I second this statement 100%

pain_fedora6 wrote:n fails/minute = average number of failed flag captures per minute

Xeno do you think that the number of failed flag captures pm is Gaussian distributed? Then the average would be representative, but, i think we should replace it with a random variable, and you should think a bit more and prepare the stochastic differential equation giving the more realistic formulation of your corollary. If we then integrate it with a nice Ito integral (Ito as the map from Red Ishad ) from 0 to timelimit we will get the total - score provided that we get reasonable statistical data for the volatilies of each players playing style.

phd training... worthless... i should remove the title chemeng phd for nexeng...

The type of distribution is a function of the team which is an independent variable and not predictable outside of a detailed analysis involving a profound application of chaos theory. This is irrelevant however because this calculation is not a predictive model for the set of all games but rather a descriptive model of a given single game, which thus eliminates the intricacies that you describe in your reply. While the distribution of fails per unit time may not conform to any standard model, the previous statement remains valid within its intended scope and provides a qualitative insight to the consequences of poorly chosen player strategies. Furthermore, this model fits our empirical data remarkably well, as well as that of others (Dave et al., 2008, esteel & divVerent, 2007).

If you are able to advance an alternative hypothesis capable of refuting this one, my colleagues and I would be very interested to review it.