Reign of Gaming

Bloodthirster snowballs better than Ravenous Hydra, so getting bloodthirster is pretty much a no brainer if your ahead. Ravenous Hydra is more of a reasonable choice if your behind in lane or breaking even against your lane opponent. It is also worth noting that Ravenous Hydra has an active that you need to use for it to be effective. If you forget to use it or time it wrong you would have been better off with bloodthirster. Ravenous Hydra users also tend to prefer a tankier build path in my experience. Bloodthirster is also a little more vulnerable to attack speed slows.

BT has a higher win rate than Hydra for one simple reason, It has a less forgiving build path, ergo people are going to be more likely to build it if they are further in front, and less likely so if they are behind, regardless of actual power when built.

Don't need maths for everything guys :)

Yes. Sorry that I mixed with other consideration. It was meant as a do not use "winrates" as a measure because it will add other factor.

The math part was that in the article it is neglected the lifesteal portion of the two objects. It says : "for now let's assume we are interested in damage only"...and then you reach the conclusion and end of the article without a "And now let's add the lifesteal component". Just that.

You've made good points. Did you get a chance to see my summation of their lifesteal contribution?

Hydra gives added value to riven because she can use it to negate cast time on w

How do you account for the fact that EVERY single champion on Lolking that shows both Bloodthirster and Hydra in their most popular items (which implies user diversity, and the viability of both items) has a higher win-rate with Bloodthirster?

This list includes Aatrox, Fiora, Kha'Zix, Master Yi, Pantheon, Riven, Talon,Tryndamere, Yasuo, and Zed.

This list does not include (since Bloodthirster does not appear in the most popular items, while Ravenous Hydra does) Darius, Gangplank, Jarvan IV, Lee Sin, Nocturne,Olaf, Renekton, Rengar, Shyvana, Sion, and Wukong.

While finally, for these melee AD champions neither option appears among their most popular items: Garen, Hecarim, Jax, Irelia, Nasus, Poppy, Skarner, Trundle, Udyr, Vi,Volibear, Warwick, and Xin Zhao.

From this information, you can see that the champions who focus on neither of these items (the third list) either have a means of sustain built into their kit, their team-fight role isn't to be a main source of damage, or they have poor AD scaling.

Champions in the second group become a bit more ambiguous. For every champion in this group, Hydra is the only item of the two that appears in their most popular items. The most obvious benefit of Hydra over BT is additional wave-clear. Champions who can utilize this well are those with AA resets/on-hit abilities, or who have low wave-clear potential in their abilities. Additionally, champions that don't suffer from poor wave-clear, but benefit from building one AD item at most (like Shyvana and Renekton) more readily choose Hydra.

The final group in question, and the one that all of this is about, is the dedicated physical damage dealers. Every single champion in this group benefits from having more than one AD item in their build, has good scaling, and is a strong offensive duelist. Everyone in this group also has a higher win-rate with Bloodthirster than they do with Ravenous Hydra.

What I'm getting at is that Hydra is nowhere near universally better, as your post would suggest. Hydra is good for champions who want to boost their wave-clear, and that is it (as evident by it's pick popularity, and lower win-rates on controversial champions). Bloodthirster is a better item for any champion that will be stacking AD (as is evident by it's higher win-rates on all applicable champions). How else would you explain the higher win-rates? I define 'better' as "Will help to win more games than the alternative choice".

I think it needs more data, more objective. His approach is mathematical, yours take win rates. Some points to clarify:

-Do people build both in the same game? in which order? If people build constantly hydra before bloodthirster, there are games you loose before having both.
-Do people build one of them when strictly ahead? Or is it better when behind? Expecially true for the "farming" element of BT. You are not incentivated to build BT when you are not in a loosing position and can loose the stacks. Hydra makes you farm faster, but sideeffect to make you push the lane in a less safe position.
-LS: this is one thing the analysis does not comprehend. BT has more LS. I would really like to have this counted in DPS inside the analysis as a sum. If kha vs. kha one with RH one with BT, who dies first?

How are data points not mathematical? Hydra + BT is HUGELY inefficient, and not a single champion in the game would have both of these in an optimal build. There are other items with compounding utility/passives to consider beyond lifesteal (like armor penetration, attack speed, cooldown reduction, crit chance, resistances, health, various item-specific passives).

The only logic for the ahead/behind argument is that Hydra has an easier build path. This may lead people to assume that they would be better off buying Hydra's pieces rather than waiting for the 1550G to afford a B.F. Sword. This mentality could also contribute to Hydra's lower win-rate, since those who start behind tend to stay behind.

The stacks of Bloodthirster are negligible, and take less than a minute to maximize after dying. Hydra can't necessarily make you farm faster unless you're jungling, since the waves won't come any faster than every 30 seconds. However, it can make you PUSH faster.

To compare the items in terms of damage output, I'll assume 2 Kha'Zix's at level 9 with base stats and having built only Bloodthirster/Hydra (the earliest point that a champion can realistically expect to finish an item of this cost), respectively. They both have 1697 Physical Effective Health (1195 HP + ((42Armor/100)1195) = 1696.9), and an Attack Speed of .816. I'll simplify this to .8 Attack Speed. Using only Auto-Attacks (since abilities would skew the results further in favor of Bloodthirster), I will include the Ravenous Hydra Active and HP/5 after I calculate the items differences in Auto-Attacks and lifesteal.

The Bloodthirster Kha'Zix has 178 AD, and 20% Lifesteal. It will take him 10 Auto-Attacks (9.5, but you can't attack .5 times), and 8 seconds (10x.8) to deal 1697 Physical Damage. He will heal 339 (1697x.2) Health in this time.

The Ravenous Hydra Kha'Zix has 153 AD, and 12% Lifesteal. It will take him 11 (11.09, but I'll simplify) Auto-Attacks , and 8.8 seconds (11x.8) to deal 1697 Physical Damage. He will heal 203 (1697x.12) Health in this time.

Now, each Kha'Zix has to re-deal the damage that was healed. This is 203 more Health for the Bloodthirster to hit through, and 339 more Health for the Hydra to hit through. The Bloodthirster will require 2 more Auto-Attacks (203/178 = 1.14, or 2) to kill his opponent, and heal 40 (203x.2) Health from these attacks. The Hydra will require 3 more Auto-Attacks (339/153 = 2.21, or 3) to kill his opponent, and heal 40 (339x.12) Health from these attacks.

In both cases, 40 health is not enough to require another auto attack, even after resistances, so we can conclude that the Bloodthirster user will need 11 (9.5+1.14) Auto-Attacks to kill his opponent through their lifesteal. The Hydra user will need 14 (11.09+2.21) Auto-Attacks to kill his opponent.

It takes Kha'Zix 8.8 seconds to attack 11 times, so there is only enough time for the Hydra user to use it's Active once. At melee range, the Hydra Active adds another Auto-Attack to the user's combo. We can remove one of the required Auto-Attacks (going from 14, down to 13) as a result. The active will heal the user for 18.36 (153x.12) Health. In 9 seconds, the Hydra will passively heal for 27 (3/s x 9) Health. The Hydra will heal for a total of 45 Health due to it's Active and Passive. This is not enough health to require another Auto-Attack from the Bloodthirster user (a total of 64 additional health would be required, as 11 - (9.5+1.14) = .36, and 178 x .36 = 64) .

In conclusion, the Bloodthirster user will win, attacking his opponent 11 times. The Hydra user doesn't deal enough damage, even with one active and the passive healing, to mitigate the health provides by the Bloodthirster user's lifesteal and increased damage. The Hydra user would need 13 Auto-Attacks to kill the Bloodthirster user, and will die by the time he uses 11 Auto-Attacks.

If they were to use abilities, the results would be MORE in Bloodthirster's favor.

...First, Hydra deals AOE damage, which means as long as you are not fighting isolated, your healing ramps up quite fast. Second, Since when does BT have 20% lifesteal? Serious? This kind of dubious info just invalidates my trust for any of your calculation. Lastly, striping one passive off a item while giving another item full passive is...NOT the way to compare stuff? Like really not?

Damn, it IS 18% isn't it, lol. I'll re-do my math to consider the full effect of both items passive effects, including Ravenous Hydra's AoE healing on 1 subsidiary target (i.e. nearby minion, or enemy champion).

To compare the items in terms of damage output, I'll assume 2 Kha'Zix's at level 9 with base stats (no runes/masteries) and having built only Bloodthirster/Hydra, respectively. They both have 1697 Physical Effective Health (1195 HP + ((42Armor/100)1195) = 1696.9), and an Attack Speed of .816. I'll simplify this to .8 Attack Speed. Using only Auto-Attacks (since abilities would skew the results further in favor of Bloodthirster), I will include the Ravenous Hydra Active and HP/5 after I calculate the items differences in Auto-Attacks and lifesteal.

The Bloodthirster Kha'Zix has 178 AD, and 18% Lifesteal. It will take him 10 Auto-Attacks (1697/178 = 9.53, but you can't attack .53 times), and 8 seconds (10x.8) to deal 1697 Physical Damage. He will heal 305 (1697x.18) Health in this time.

The Ravenous Hydra Kha'Zix has 153 AD, and 19.2% (12% base + 7.2% (12x.6) from nearby enemy) Lifesteal. It will take him 12 (1697/153 = 11.09) Auto-Attacks , and 8.8 seconds (11x.8) to deal 1697 Physical Damage. He will heal 325 (1697x.192) Health in this time.

Now, each Kha'Zix has to re-deal the damage that was healed. This is 325 more Health for the Bloodthirster to hit through, and 305 more Health for the Hydra to hit through. The Bloodthirster will require 2 more Auto-Attacks (325/178 = 1.82, or 2) to kill his opponent, and heal 58.5 (325x.18) Health from these attacks. The Hydra will also require 2 more Auto-Attacks (305/153 = 1.99, or 2) to kill his opponent, and heal 58.56 (305x.192) Health from these attacks.

In both cases, 59 health is not enough to require another Auto Attack. The Kha'Zix using Bloodthirster would be able to kill his opponent in 12 Auto Attacks (9.53+1.82 = 11.35, or 12), while the Hydra user would need 14 Auto Attacks (11.09+1.99 = 13.08, or 14).

It takes Kha'Zix 9.6 seconds to attack 12 times, so there is only enough time for the Hydra user to use it's Active once. At melee range, the Hydra Active adds another Auto-Attack to the user's combo. We can remove one of the required Auto-Attacks (going from 14, down to 13) as a result. The active will heal the user for 29.3 (153x.192) Health. In 9 seconds, the Hydra will passively heal for 27 (3/s x 9) Health. The Hydra will heal for a total of 56 Health due to it's Active and Passive. When combined with the health obtained from mitigating the Bloodthirster's lifesteal, this is enough health to require another auto attack from the Bloodthirster user (59+56 = 115, and 12 AA's - 11.32 = .64, and 178 x .36 = 113.92) .

This means that the Bloodthirster user would be able to kill their opponent in 13 Auto-Attacks, AND the Hydra user would also be able to do the same. It's a draw, down to the last HP. I was not expecting that, though it is assuming the best case scenarios, in that your subsidiary target for the Hydra splash lifesteal doesn't die in 13 Auto-Attacks, and you have Maxed your Bloodthirster stacks. Using abilities would skew this result into Bloodthirster's favor.

Better? I've abandoned my argument as to which is a 'better' item, and I agree that the choice should be situational depending on whether or not you are ahead.

Yes good calculations. I will have simply added 18% to BT and 12% to Hydra in the article calculations(with simply no armor/HP because it is the same for the 2 items). Also 9.6 sec...it needs very little to have another active from hydra :D.

I think that is really "close" as a confrontation. As a general rule if you go single offensive item Hydra is better, but with atkspeed or cdr the balance will shift to BT (hydra active do not scale) and if you get hydra is because you need burst instead of dps. This considering a single target focus. On multi target hydra will be better for overall damage.

Using the same logic in which you believe that I am assuming conclusions, I think it is unwise to assume that Ravenous Hydra and Bloodthirster should ever be used in tandem. If you have Bloodthirster, there is a list of 10 items or so that will easily fill your inventory spaces more efficiently than including a Hydra.

I think that a large portion of the disparity lies in the fact that many people believe Hydra to be a good "behind" purchase because of it's efficient build path, while Bloodthirster a good "ahead" purchase (on all applicable champions who show both items in their most popular choices). It is not because these champions commonly buy one before the other, it is because these champions commonly choose the item with the simpler build path when they are not excelling in lane. And if a champion doesn't excel in lane, they will have a harder time winning their game, thus the lower win-rate.

Hydra is NOT universally better as you are suggesting, especially for champions who have no trouble whatsoever with waveclear, like Talon (seriously, wth is up with this reference?).

Fiora has the highest Hydra purchase rate, as many people believe the item to be core on her since she has trouble waveclearing even while using abilities (and it synergizes with her ultimate).

Ultimately, confusing the apex for the status quo is the only thing that I see wrong with your logic. I agree that a late-game prioritization may be contributing to the win-rate disparity, but I do not believe that Hydra is popular for the correct reasons.

Where does Bork fit in with these two?  Yes it has less raw damage (worse for skills), but there's also the AS boost as well as the active and passive it has.

What about for auto attacks?
It isn't all that difficult to figure out whether or not bloodthirster would provide more damage than hydra on auto attacks.

ad = attack damage
AS = average number of attacks made per second
CM = average damage modifier due to critical strikes

hydra(ad, as, cm) < bloodthirster(ad, as, cm)
(ad + 75) * AS * CM + ad * 0.1 < (ad + 100) * AS * CM
ad < 10 * ( (ad + 100) - (ad + 75) ) * AS * CM
ad < 250 * AS * CM

As long as your attack damage is less than 250 * AS * CM, then the additional hydra active every 10 seconds won't overcome the damage bonus provided by the additional 25 attack damage of bloodthirster, in a pure sustained damage circumstance for auto attack damage. That inequality only takes into account auto attacks, and nothing else-- no abilities.

CC = critical strike chance, as a decimal
CD  = bonus critical strike damage, as a decimal
CM = 1 + CC*(1 + CD)

for example, with phantom dancer and infinity edge, you have 55% crit chance and 50% bonus crit damage, so you have:

CM = 1 + 0.55 * (1 + 0.5)
CM = 1.825

If you also have 1.5 attacks per second you can figure that you would need:
ad < 250 * 1.5 * 1.825
You'd need attack damage less than 684 in order for bloodthirster to provide more damage than hydra on auto attacks with just phantom dancer and infinity edge. Bloodthirster is much better for melee carries than hydra is for raw single target damage. The difference provided by the hydra active for burst damage is very very quickly overcome by the 25 bonus ad being multiplied by crits applied nearly twice per second.

this line of thinking seems more plausible and usable than the one depicted in the article.

I'm not likely to buy either on melee carries though, personally. The benefits of BotRK for actually getting uptime on targets in the first place are too good to pass up.