Flag-Locking: Don't waste your money » Forum Post by lukkas

To Lieu- valiant effort. I couldn't resist taking a stab at it myself

I think that people who are arguing against the 32% swing are failing to understand some main concepts, so let me see if I can provide some insight.

1. The flag rarely, if ever, is neutral. This means that at almost any point in the game, one of side will have control of the flag and receive the buff. Because of this, one needs to take into consideration the loss of the buff by one side and the gain of the buff by the other side when calculating the advantage of completely capturing a flag.

2. Yes, one side loses the 15% buff and the other side gains the 15% buff, so when you capture a flag, you only gain 15% of your health, while the other side is returned to baseline, but Lieu- is talking about the relative change in relative strength, somewhat of a confusing idea. This is one way of looking at things, you could also talk about the absolute change in relative strength.

Example 1:

suppose TB has T_0=1410HP (lvl 1, no items) and Rook (lvl 1, no items), who has control of the flag has R_1=2185HP. TB's relative strength to Rook is

a) str_0 = T_0/R_1 = 1410/2185 = 64.53%

Now, suppose that TB makes the flag neutral. Rook loses his 15% bonus, Rook goes down to his baseline health (call it R_0). we have

b ) 2185 = 1.15*R_0 ==> R_0 = 1900

Next, suppose that TB remains near the flag, and completes the capture. TB (and his team) gain the 15% bonus. TB's new health is given by

c) T_1 = 1.15*T_0 = 1.15*1410 = 1621.5

Now TB's relative strength to Rook is given by

d) str_1 = T_1/R_0 = 1621.5/1900 = 85.34%

So the absolute change in relative strength is

e) Abs_1 = str_1-str_0 = 85.34% - 64.53% = 20.81%

and the relative change in relative strength is

f) Rel_1 = str_1/str_0 = 85.34%/64.53% = 132.24% = 100% + 32.24%

Example 2:

suppose TB has T_0=3500HP (lvl 20, no items) and Sedna (lvl 20, no items), who has control of the flag has S_1=4352.75HP. Then TB's relative strength to Sedna is

a) str_0 = T_0/S_1 = 3500/4352.75 = 80.41%

Now, suppose that TB makes the flag neutral. Sedna loses his 15% bonus, Sedna goes down to her baseline health (call it S_0). we have

b ) 4352.75 = 1.15*S_0 ==> S_0 = 3785

Next, suppose that TB remains near the flag, and completes the capture. TB (and his team) gain the 15% bonus. TB's new health is given by

c) T_1 = 1.15*T_0 = 1.15*3500 = 4025

Now TB's relative strength to Sedna is given by

d) str_1 = T_1/S_0 = 4025/3785 = 106.34%

So the absolute change in relative strength is

e) Abs_2 = str_1-str_0 = 106.34% - 80.41% = 25.93%

and the relative change in relative strength is

f) Rel_2 = str_1/str_0 = 106.34%/80.41% = 132.25% = 100% + 32.25%

So as you can see, the absolute change in relative strength depends on the exact numbers, while the relative change in relative strength is a constant, and therefore a very useful number. It's useful because it gives you an easy way to compare your HP to another hero's HP before and after the flag capture. The 15% is only useful for comparing your HP to another's if you have a calculator with you.

Here's how to use the 32% in practice:

Suppose TB has about 3/4 (75%) of the health of Sedna, who controls the health flag. TB knows that if he captures the health flag, they will have the about the same health. Why? because he knows that his relative strength will increase by a factor of 32% (about a third). One third of 75 is 25, and 75+25=100. His final relative strength to Sedna will be about 100%, meaning they both have the same HP.

This tells us that the higher your relative strength to the enemy, the more you stand to gain if you captures the flag, and the more you stand to lose if your enemy captures the flag.

Conclusion: Flags are important.

Edit: For those of you who may doubt these numbers, let me say that I am currently working toward a Ph.D. in math, and was a math major in undergrad. I solve many math problems on a daily basis. I also tutor high school student in math regularly, so I know the high school curriculum in and out, and I can tell you that this problem has about an 8th grade level (give or take a year or two, depending on the student's ability). If you disagree with the numbers, then I suggest you reread my post, because there is something you don't understand. If you read it a second time and still don't agree with the numbers, please feel free to PM me with whatever your complaints are.

+1

Successfully updated karma reason!

Wow, the numbers get really their own lives here i think.

Here is another meaning:

If i hold a flag and gain a 15% bonus on my opponent then this is the objective advantage i have. Then i lose the flag and in the next fight my opponent has a 15% advantage over me. Then the objective advantage of him over me is this 15%. My subjective disadvantage lies at around 30%, because i lost my 15% and he got it.

Alex

Successfully updated karma reason!

The point is that I understand that, that is not my problem.

It is the two different sides of the same coin.

The problem I have is that it can be misleading.

Yes, the status change is 32%, however at the same time you are still only 15% more powerful than your opponent. You are not 32% more powerful, you are 32% more powerful than the situation you were in before. However, in real terms you are only 15% more powerful than your opponent who is at base. This is just as valid as most of the other damage/HP objects (tower/creeps in the game will not be affected by the change).

No one can argue that capturing the flag is not important, whether you are in a swing state of 32% better off or a real state of being 15% more powerful, makes no bloody difference, they are both better than having less HP.

Whether it is worth locking the flag over it, is another question. Not always no, especially if the flag is something like -15% cooldown.

Successfully updated karma reason!

Yes, the status change is 32%, however at the same time you are still only 15% more powerful than your opponent. You are not 32% more powerful, you are 32% more powerful than the situation you were in before. However, in real terms you are only 15% more powerful than your opponent who is at base. This is just as valid as most of the other damage/HP objects (tower/creeps in the game will not be affected by the change).

That would be only correct if the opponent was 15% more powerful that you when he had the flag and this is extremely rare condition as there are too many other factors into play. That aboslute "15% more powerful than the opponent" is a misleading single case. What happens if you was already weaker than the opponent before he took the flag. Or if you were stronger even after he took the flag. Then you wouldn't be absolutly 15% more powerful than him after flag exchange.

The easiest way as I already said is as follows, Demigods A and B are going against eachother with A owning the flag. Doesn't matter who is more powerful. B can do the following things with pretty much the same end result:

Take the hp flag
Buy 32% more hp
Buy 32% more dps (assuming no abilities are used)

The ending relation in their power will be the same regardless of which one of the above happens. This very likely will not result in B being exactly 15% more powerful than A. The 32% item = taking flag is the correct way to think about this, nothing else.

Successfully updated karma reason!

Quoting Sinzer,

reply 3

The point is that I understand that, that is not my problem.

It is the two different sides of the same coin.

The problem I have is that it can be misleading.

Yes, the status change is 32%, however at the same time you are still only 15% more powerful than your opponent. You are not 32% more powerful, you are 32% more powerful than the situation you were in before. However, in real terms you are only 15% more powerful than your opponent who is at base. This is just as valid as most of the other damage/HP objects (tower/creeps in the game will not be affected by the change).

No one can argue that capturing the flag is not important, whether you are in a swing state of 32% better off or a real state of being 15% more powerful, makes no bloody difference, they are both better than having less HP.

Whether it is worth locking the flag over it, is another question. Not always no, especially if the flag is something like -15% cooldown.

The flag having a value of 32% is completely "real". 15% is just your final state, which of course you consider when making decisions, but the 32% difference is absolutely physical and matters a huge amount when it comes to evaluating the usefulness of using flag locks on that flag.

Stop trying to blow off the usefulness of the 32% value. Both 15% and 32% are different pieces of information and they are both useful.

Successfully updated karma reason!

Agreed, both percentages are useful. Which one is the more useful way of considering the situation depends on what you're evaluating.

Suppose you're deciding between two actions that will give you health: (1) running back to the item shop to buy a piece of health armor or (2) running across screen to capture a health flag that your opponent has. Suppose further that both actions grant you the same amount of absolute health (+15%), and gold isn't an issue.

If you think you're going to engage your opponent in autocombat soon and want to maximize your chance of killing him 1v1, the 32% value is the more useful number because it best captures the relative change of your health versus his that results from taking the flag. When fighting 1v1, it's a race to who gets the other guy to 0 health first, and any action that boosts your hp while reducing your opponent's is obviously very advantageous for you. Consequently, you might decide to expend the effort to capture the health flag.

If on the other hand, you just want some more snipe or venom spit insurance because you are torch bearer and you aren't going to really engage 1v1 melee, but are getting pushed back because someone keeps venom spitting you, then you might decide that all you really care about is getting +15% in the fastest, safest way possible. From that perspective, getting the armor results in the same benefit as getting the health flag, and is a more efficient way to accomplish your goal.

Successfully updated karma reason!

Quoting Jomungur,

reply 6

Agreed, both percentages are useful. Which one is the more useful way of considering the situation depends on what you're evaluating.

Suppose you're deciding between two actions that will give you health: (1) running back to the item shop to buy a piece of health armor or (2) running across screen to capture a health flag that your opponent has. Suppose further that both actions grant you the same amount of absolute health (+15%), and gold isn't an issue.

If you think you're going to engage your opponent in autocombat soon and want to maximize your chance of killing him 1v1, the 32% value is the more useful number because it best captures the relative change of your health versus his that results from taking the flag. When fighting 1v1, it's a race to who gets the other guy to 0 health first, and any action that boosts your hp while reducing your opponent's is obviously very advantageous for you. Consequently, you might decide to expend the effort to capture the health flag.

If on the other hand, you just want some more snipe or venom spit insurance because you are torch bearer and you aren't going to really engage 1v1 melee, but are getting pushed back because someone keeps venom spitting you, then you might decide that all you really care about is getting +15% in the fastest, safest way possible. From that perspective, getting the armor results in the same benefit as getting the health flag, and is a more efficient way to accomplish your goal.

Excellent point!

Successfully updated karma reason!

Quoting Misfortune,

reply 4

The easiest way as I already said is as follows, Demigods A and B are going against eachother with A owning the flag. Doesn't matter who is more powerful. B can do the following things with pretty much the same end result:

Take the hp flag
Buy 32% more hp
Buy 32% more dps (assuming no abilities are used)

The ending relation in their power will be the same regardless of which one of the above happens. This very likely will not result in B being exactly 15% more powerful than A. The 32% item = taking flag is the correct way to think about this, nothing else.

I understand where the 32% difference comes from, but I think it is taken a bit out of context in a lot of these examples. In this scenario, you would only need 30% more HP or 30% more DPS to make things equivalent.

The 32% difference is from the enemy's perspective after the flag has been taken. They would need 32% more health than they have now to attain the same proportional amount of HP as before. Looking at it from the other side, they would need to lose 24% of their health while they have the flag to be at the same proportion they would be at after you took the flag. It is difficult (either way) trying to look at this from the enemy's perspective.

The problem with all this theory is that two demigods rarely ever have the same health. If you're at 5,000HP and I'm at 3,450HP (with the flag), we can't measure the "battle advantage" proportionately. Now, when I lose the flag, it is a 1,200HP swing in your favor. It is still the 32.25% gain (proportionately), but now measuring it in that manner doesn't reflect your actual advantage in a fight.

Your actual advantage of gaining the flag was from ~45% HP more than me to ~91% more HP than me (relative to my health). On the flip side, when I take the flag back, I go from having ~52% of your health to ~69% of your health (obviously relative to your health).

As obvious as it sounds, the larger the health difference between the teams, the more this percentage skews and the more important the health flag becomes.

Successfully updated karma reason!

Quoting Milskidasith,

reply 25

Quoting RobWild,
reply 24
After the start of the game somone almost always has the health flag. When you take it back

You gain 15%

They lose 15%

Therefor your advantage is 30%, I think people are confused by the actual real percentage of 32.5 or what not adding up, just read the first three lines of my post again, its all you need to know.

It's not 30%. The math actually works out to be 1.15^2 advantage, not 1.15X2 advantage.

I totally understand that Mis, re-read my post. I was just trying to simplify it, I think you guys are losing everyone throwing around the 32 number even though it is mathmatically correct. Its simpler to just say +15 for you -15 for them results in a 30% advantage in health for you.

Successfully updated karma reason!

If you're at 5,000HP and I'm at 3,450HP (with the flag), we can't measure the "battle advantage" proportionately. Now, when I lose the flag, it is a 1,200HP swing in your favor. It is still the 32.25% gain (proportionately), but now measuring it in that manner doesn't reflect your actual advantage in a fight.

Exactly That is why I am saying it is misleading, not that it is false.

Successfully updated karma reason!

Flag benefits aside, another use for locking flags early are teleport kills. On the intial rush to cap flags, if you get control before or during a 1v1 encounter at a flag, lock it and run back to the crystal to regen. Most likely your level 1 opponent will still hang around to farm creep even though the flag is locked and hes at 50% or so hp because he wants to get started leveling up. If you are efficient enough in getting back to the crystal to regen, you can port back and most likely get the kill while hes working on creep. This only tends to work on the initial push though when people are level 1-2. Even if there are 2 people at the flag, if there are no healers and one is low, you had a good chance of finishing him off and getting away. Levithan is a perfect map for this as you have a nearby tower to retreat to that will help you escape.

Also keep in mind a fair number of people have the graphics turned down. Unless fedelity is high, the teleport graphics are not as obvious so there is a high chance of them not even knowing you are porting in.

Successfully updated karma reason!

This whole thread is kinda funny yet painful at the same time. Basically you have two different arguements here. One side is tallking about the present with no reference to history or what happend before right now. Meaning, I have the flag therefore my advantage right now is 15%. The other side is taking the past into account, saying that you had the flag and I took it, so you went from +15% to even and I went from even to +15% therefore the change from then to now is a total of 30%(keeping it simple) So technically both sides are right.

Take a much easier example. I have $10 and you have none so I have $10 more than you. If I give you the $10 you now have $10 more than me but the total difference, from before to after the exchange, is $20

Successfully updated karma reason!

Taking a flag causes +15% hp for one side and -15% for the other pretty much simultanously in 99% of the cases. We're friggin only trying to measure these 2 changes to determine the value of locking a flag. There is no 15% advantage gain from taking the flag (though you could make an argument about 15% HP INCREASE!!!). Advantage is a relative term...

Successfully updated karma reason!

The confusion comes in because you would use different numbers depending on the decision (as Jomungur pointed out):

Should I engage my enemy? (Take the current, visible 15% advantage into account.)

Should I take and/or lock the flag? (Take the ~32% change in advantage into account.)

The math is undebatable. The conversation is about how you use the numbers.

IMO health, mana, and celerity flags are almost never worth locking. Portal, gold, Xp, and at times valor flags are worth it. XP/gold earlier on, and portal/valor later in the game when grunts become important.

Artifact flags are in a league of its own, as it's highly dependent on the situation. If both sides are getting high gold, then it's definitely a clincher to keep it locked. Never had to do that, though.

Successfully updated karma reason!

I ain't ever going to capture a flag again

All this maths makes my head hurt

I just want to kill

Successfully updated karma reason!

Well. Gotta say, this has been very interesting. I was ready to abandon my position, clad in a cloak of shame.

But the internets, ever vigilant for the travails of the weak and just, have exonerated me.

Is 32% the right number? Of course.

What about 15%? Is that relevant? Highly relevant.

I stand before you, chastened for my overly zealous defense of the number 15 but exalted for that same defense.

Shadow_Avenger, standing on the shoulders of others who came before, encapsulated the salient points perfectly:

The confusion comes in because you would use different numbers depending on the decision (as Jomungur pointed out):

Should I engage my enemy? (Take the current, visible 15% advantage into account.)

Should I take and/or lock the flag? (Take the ~32% change in advantage into account.)

I propose that, after 5 pages of courageous debate and mind-numbing repitition, we close the discussion.

Successfully updated karma reason!

I don't think there is ever a question of whether or not you should take the flag.

Successfully updated karma reason!

Quoting Milskidasith,

reply 17

I don't think there is ever a question of whether or not you should take the flag.

Well, it's opportunity lost while you could be doing other things. BUT, I suppose the issue is more about locking and/or deciding which flags to keep, etc.

Successfully updated karma reason!

I propose that, after 5 pages of courageous debate and mind-numbing repitition, we close the discussion.

Amen to that!

Successfully updated karma reason!

Actually, as value has been settled, we can now discuss the OP - are locs worthwhile

Successfully updated karma reason!

Quoting Misfortune,

reply 20

Actually, as value has been settled, we can now discuss the OP - are locs worthwhile

Yes they are! Situationally

Successfully updated karma reason!

if yur oppnent iz dum than u can lok flag and he will just not no wut 2 do it works gud imo

Successfully updated karma reason!

I think what the 15-people are trying to say is that you still have only 15-percent extra hp. While you did get an 32.5 percent advantage mathematically.

However you arent suddenly steamrolling your opponent because you got 32.5% extra hp, you changed the balance from you having an slight disadvantage to you having aslight advantage. Now, mathematically this 32.5% is totally correct. But in game this doesnt mean you are suddenly much stronger than your opponent. For some less intelligent people this may be a bit misleading.

For the rest... no remorse for discalculic morons.

Successfully updated karma reason!

Quoting transitive,

reply 22

if yur oppnent iz dum than u can lok flag and he will just not no wut 2 do it works gud imo

Unless they are so dumb that flag-locking becomes a waste of money.

Successfully updated karma reason!

Sorry that I can't be arsed to read this whole thread but just wanted to point out that almost everything written here so far is wrong. If I flaglock the xp flag on catarak it will give my team more xp. If I do not flaglock it will not nessesarily mean that we won't have the flag anyway. If I flaglock the xp flag the opponents will not stand there. They will put more pressure on another flag instead. Assuming that the opponents are stronger in combat and would have taken the xp flag they will probably succeed in taking the 15% hp flag instead and they can also lock flags. If I flaglock a flag I don't have to stay there and protect it but the opponents does not have to be there either.

I can not say if flaglock is worth the money or not but I can say this: Neither do you. I would guess that the power of flaglock are overestimated.

This problem can not be solved with the use of numbers. This is a logical problem and it is extremly complicated.

It is an intresting discussion but we need to think outside the current box. Attack this problem from another point of veiw.

Successfully updated karma reason!