A Strictly Economic View of Shield Use in PvP

helix72

I was thinking today and wondering to myself...when is it worth it to spend HP to shield in PvP, and how much is worth spending? Or by shielding am I actually burning more resource than I'm winning back? So I set out to answer that question.

I've built a model that effectively converts every asset into an expected number of 5* covers. So first, some disclaimers:

1) This treats every 5* as equally valuable. We all know that isn't the case, but we have to start somewhere. The model I ended up building actually allows me to value 5s differently, but not everyone agrees on where they rank, so for now in my personal model I've devalued any 5 I already have at 550. That is not included below.
2) The conversion rate of HP to 5* covers significantly influences the answers. Based on my actual returns in the vaults over the last few months, I've assumed a net conversion rate of one 5* cover for every 1250 HP. Before you think that sounds low, keep in mind this is a net rate, meaning it is the amount I spent to draw from the vaults minus the amount I earned back from the draws and any other rewards.
3) You don't always have the option to buy the 5* you want with HP! This is a big point, if you don't have a wide array of 5's you're trying to level the fact that you can buy them for HP is irrelevant if the one(s) you want to buy are never offered.
4) The asset values change, even if only slightly, with every new character that is introduced and any feeder changes. I recently updated for the new characters through and including Abomination and the next 5* and updated all the characters for the feeder updates.
5) This only concerns placement and assumes you hit progression without the use of shields.
6) This ignores other uses of HP. I do realize that HP is used for boosting and even some people still use it to buy healthpacks when they run out and either don't have the tokens or are too lazy to pull taco tokens to get more without spending (guilty!).

So with the above in mind, using the current award structure for the Hulkling PvP (but not the specific characters awarded), here is what I came up with as the effective amount of net HP earned by Placement in SCL 10:

So what can be done with this information? You can find the point in the placement where you maximize your return. If you can get t50 by climbing in the last 3 hours and using a single 3 hour shield, your expected net return is 691-75=616 HP. In contrast, if you spend a 3, an 8, and a 3 to get t25, your net return is only 896-75-150-75=596, in the long run a worse result.

As mentioned in the disclaimers, this doesn't account for differences in the value of 5* characters, but you can do that on your own. Maybe for your own play you decide if the 5* reward is Wasp you don't care about placement at, just about maximizing your net HP, and if the 5* is Shang Chi you don't care about maximizing your net HP, you just want the most shards possible. That's entirely up to you, but at least now you may have an idea of how much that extra shielding is really costing you in the long run.

As always, constructive feedback and sincere questions are welcome!

entrailbucket

Ah, but one simply cannot put a value on the feelings of pride generated by seeing your name in 1st place with 5,000 points!

...seriously though, more players should think about things this way. I'm not sure I agree with the value you've got on 5* shards (it does feel low to me), but I also can't come up with a better conversion rate.

atomzed

Ooh I would like to see your model! What did you use to develop the economical model?

I like how you summarise your results without going into the details. For a lay player this is a good rule of thumb.

helix72

@atomzed i started by putting every character in the game in a spreadsheet and then mapping out each champ reward for all of them. For a 5*, for example, they need 13 covers to champ and then have 100 champ levels, so I took the total rewards and divided by 113 to get the “per cover” rewards.

Then I needed to convert every asset in the game into 5* covers. It requires some iterative calculations as, for example, a 4* champ reward can be an LT, but an LT can give you a 4*.

Since the token odds are published, it was fairly straightforward to plug the formulas in and just calc until it converged. I did need to distinguish which characters can be found in which token stores, removing limited 4*s for example.

Tomorrow if folks are interested I can post some of the current values. Note I give 0 value to ISO, RISO, and 1* characters in the model but everything else has a value. For 3* and 2* I also include the value of the HP from selling a max champed version. I’m nowhere close to selling max champed 4* or 5* so I didn’t include that HP, but I set the model up so I could.

entrailbucket

The model (or, well, "model," since I ain't doing all that math) I always wanted would convert all the rewards into cash. I know everyone is free to play now, but back in the day people spent a lot of money on shields, and I always thought they'd be better off just buying those rewards outright.

atomzed

@entrailbucket said:
The model (or, well, "model," since I ain't doing all that math) I always wanted would convert all the rewards into cash. I know everyone is free to play now, but back in the day people spent a lot of money on shields, and I always thought they'd be better off just buying those rewards outright.

Actually, I think helix model would be able to do that. If I understand his description correctly, he converts all the rewards into HP. Example he converts the value of 5* shards into HP. That makes it possible for him to analyse how much value (in terms of HP) one would receive for placement.

With the HP, it’s easy to convert that into cash value. Just the standard purchase of HP with cash.

entrailbucket

If that works, what'd be really fun is working out a theoretical hourly rate for your MPQ "labor."

Imagine some horrifically unfun event that you're playing only for rewards -- you could figure out that the game is effectively "paying" you something like $1.50/hr and spend your time on a more lucrative event with a higher pay rate.

atomzed

@entrailbucket said:
If that works, what'd be really fun is working out a theoretical hourly rate for your MPQ "labor."

Imagine some horrifically unfun event that you're playing only for rewards -- you could figure out that the game is effectively "paying" you something like $1.50/hr and spend your time on a more lucrative event with a higher pay rate.

Yup! And it’s also possible to convert the current roster to a dollar value. Ie how much would your 5* and all characters be worth.

Sekilicious

I like this. I’m not entirely sure how easy it is to push into the top-50 with 3hrs left though. Depends on the slice maybe. I do know that nearly all are shielded in Slice 1 even with 5 hours left. Except the people running Chasm + either iShulk/mThor/iHulk, of course. Part of the value of hopping is that there are more targets available that are worth more.

atomzed

@helix72 said:
2) The conversion rate of HP to 5* covers significantly influences the answers. Based on my actual returns in the vaults over the last few months, I've assumed a net conversion rate of one 5* cover for every 1250 HP. Before you think that sounds low, keep in mind this is a net rate, meaning it is the amount I spent to draw from the vaults minus the amount I earned back from the draws and any other rewards.

Helix, can you explain more about your assumption of 1250 HP = 5* cover?

This seems awfully low. Let me use the official offer as an illustration of the worth of a 5* cover.

1) In Daily Shard Offers, 20x 5* shards is a the lowest price of 560 HP. We know that the price increase with each purchase, but for illustration purpose, let’s assume it stays the same. To get 1x 5* cover, we need to purchase the lowest offer 25 times. Which cost a conservative price of 14,000 HP. This is 10x more than your estimate.

2) We can purchase hero shards directly with cash. The cheapest I have seen is 1050 shards for around 45 USD dollars. How much HP would you be able to get if you have paid 45 USD? Again, using the in game store of a Stark Salary, each USD gives you around 200 HP. That means 45 USD could cost you around 200 HP. Which means the 1050 shards would have cost you an equivalent of 9000 HP. Divide it into half, a 5* cover would be around 4500 HP.

3) One can buy a 5* cover with 250 CP. how much would 250 CP cost? Assuming you pay for a Stark in a buy club, you get 140 CP. You would need around 2x Stark to be able to afford a 5* cover outright. That is around 40,000 HP.

We can debate over which one has more value-for-money. I certainly not intending to debate on this, since I think the typical offers are over-priced (cough, iso offer, cough). But the 3 scenarios aim to demonstrate that your assumption of 1250 HP = 1x 5* cover is really low.

I really like to hear your steps in deriving that ‘exchange’ rate! It sounds like a fun academic exercise.

P.S. I not sure about the actual pricing of Starks in USD because my pricing is based on my local currency. I only did a rough estimate, using the exchange rates.

atomzed

@atomzed said:

@helix72 said:
2) The conversion rate of HP to 5* covers significantly influences the answers. Based on my actual returns in the vaults over the last few months, I've assumed a net conversion rate of one 5* cover for every 1250 HP. Before you think that sounds low, keep in mind this is a net rate, meaning it is the amount I spent to draw from the vaults minus the amount I earned back from the draws and any other rewards.

Helix, can you explain more about your assumption of 1250 HP = 5* cover?

This seems awfully low. Let me use the official offer as an illustration of the worth of a 5* cover.

1) In Daily Shard Offers, 20x 5* shards is a the lowest price of 560 HP. We know that the price increase with each purchase, but for illustration purpose, let’s assume it stays the same. To get 1x 5* cover, we need to purchase the lowest offer 25 times. Which cost a conservative price of 14,000 HP. This is 10x more than your estimate.

2) We can purchase hero shards directly with cash. The cheapest I have seen is 1050 shards for around 45 USD dollars. How much HP would you be able to get if you have paid 45 USD? Again, using the in game store of a Stark Salary, each USD gives you around 200 HP. That means 45 USD could cost you around 200 HP. Which means the 1050 shards would have cost you an equivalent of 9000 HP. Divide it into half, a 5* cover would be around 4500 HP.

3) One can buy a 5* cover with 250 CP. how much would 250 CP cost? Assuming you pay for a Stark in a buy club, you get 140 CP. You would need around 2x Stark to be able to afford a 5* cover outright. That is around 40,000 HP.

We can debate over which one has more value-for-money. I certainly not intending to debate on this, since I think the typical offers are over-priced (cough, iso offer, cough). But the 3 scenarios aim to demonstrate that your assumption of 1250 HP = 1x 5* cover is really low.

I really like to hear your steps in deriving that ‘exchange’ rate! It sounds like a fun academic exercise.

P.S. I not sure about the actual pricing of Starks in USD because my pricing is based on my local currency. I only did a rough estimate, using the exchange rates.

For completeness sake, there’s a 4th scenario which I didn’t mention. We can buy a 5* cover with 720 CP. That is 2.88x more than the 3rd scenario (of using 250 CP). Which is equivalent of around 115,200 HP. That is just a terrible exchange rate. Not worth it.

HoundofShadow

I calculated that 1cp = 100 hp using the difference between the first two whales costumes a few years back.

Thor Costume Bundle (74.99)
Contents:
15,000 Hero Points
75 Command Points
Thor (Gladiator) - Thunder Costume

Red Goblin Bundle (74.99)
Contents:
10,000 Hero Points
5x Masked Vengeance Tokens
Red Goblin Costume

The difference is that 75 cp and 5000 HP has been traded for 5 CP store tokens, which is worth 125 cps.

5000 HP = 125-75 = 50cps
Therefore, 1 CP = 100 HP

This is almost in line with the cost of getting a full 5* cover from shard store.

Here's the test:

1) 5* HP store
Odds is 1:166 for 40x, which costs 11800 HP. 4x 11800 = 47,200 HP.

2) 1 cover worth of 5* shards cost in Shard store
Cheapest: 14000

Most Expensive: 6395 + 1200*12 = 20795 + 3415 = 24210

3) HFH 5* Cost
Classic cover = 250 cp in HfH which is about 25,000 HP
Latest = 50,000 HP
Latest 5* HP store = 1:166

The difference between 14000 and 24210 or 250cp and 720cp is how patient the players are. That's why it costs 720 cps to buy unlocked power instantly without RNG and you can choose the colour, assuming all three powers are unlocked, compared to HFH or random pulls. That expensive option is for impatient players. The difference in cost can be boiled down to how patient and how risk averse the players are. That's the premium players pay for these.

However, since BCS has effectively devalued 5* covers, 1 CP isn't worth 100 HP anymore.

helix72

@atomzed Here's an example of how I track the HP per 5* cover:

The last holiday vaults that were offered featured iHulk and Colossus. The iHulk vault was a much better value because it would give you back 1500 HP and only cost 5500 HP (via the daily deal) to buy 40/44 items. The daily deal was 60 for the single and 1600 for a 10-pull.

I ended up doing 42 pulls spanning 2 days, so I spent 5500 + 60 + 60 = 5620, here's what I got, and what the value of those were:

In case you're wondering, I pulled 42 because I pulled a 40 pack but didn't get all the iHulk shards, so just used the daily single twice until I got them all.

So I spent 5620 HP, immediately got 1500 back (two 500 HP awards and the dupe Mysterio costume was worth 500 HP also), and had another 1325 in expected HP from the other awards, for a net spend of 2795 HP. For that net spend, I got 3.83 "expected" 5* covers, so for this vault pull round it was 729 HP per expected 5* cover.

Not all HP deals are created equal, the vaults we saw at year end were much better than average which is why I pulled almost every daily deal I could, but also why I'm using a conversion number (1250) that is higher than what I actually got in the YE vaults. I wanted some Colossus and I pulled some from that vault as well, unfortunately I didn't save my pulls but I do remember my net spend per 5* was significantly more because it didn't have the 1500 HP return, it had Spider-Ham covers which were worth much, much less.

Bad

The calculations could be right if the game was played individually.
However as we know the points scored on pvp might serve for the alliance ranking higher or lower, and that can suppose a high increase on HP or not, aside the 4* covers and 5* shards (and in the same way a 4* cover can mean a lot depending on the cycle which is farming).
Everything is interconnected and calculations get more difficult.

fractalvisions

Nice way of looking at it!

I think there's something you've missed though, which is to consider how high you'd end up if you didn't use any shields.

If you can get t50 by climbing in the last 3 hours and using a single 3 hour shield, your expected net return is 691-75=616 HP.

Sometimes the 3 hour shield can mean a difference between t50 and t200, but sometimes you can get t100 even without a shield. By your accounting that would mean the shield is a net return of 691-558-75 = 58 HP. Fortunately still positive but perhaps only if you rate the 5* highly and/or don't have much confidence in how defensive your team is.

entrailbucket

I think that's why that number for 5* shards feels low to me -- it's based on probably the best deal they've ever offered. I'm not sure what the right number is, though, and I suppose that's as good a choice as anything else. Maybe it would feel closer if we knew that deal was going to come back more regularly.

atomzed

@helix72 said:
@atomzed Here's an example of how I track the HP per 5* cover:

The last holiday vaults that were offered featured iHulk and Colossus. The iHulk vault was a much better value because it would give you back 1500 HP and only cost 5500 HP (via the daily deal) to buy 40/44 items. The daily deal was 60 for the single and 1600 for a 10-pull.

No wonder! The holiday vault was really the best deal, ever! I actually spend a lot of HP on it, pulling for apocalypse. So much so that I refresh the vault 3x.

As it is such a great deal, it significantly lower the value of a 5* cover. You could technically double the cost, and at 2500 HP per cover (which falls in line with Hound estimate of 1 CP = 100 HP), it would still be considered conservative.

Nevertheless, this doesn’t detract your conclusion. If you have the model on google spreadsheet, you could just allow the user to specify the value of the 5* cover, and let them see the change in value.

helix72

Keep in mind that was just one vault, and my net from that vault was significantly better than the conversion rate I’m using (729 vs 1250). The conversion rate I’m using is based on a longer term average, which, granted, also included anniversary. The question is: is this a trend with our new benevolent overlords or a one time thing?

Regardless, I built the model so I can put in different conversion rates. Throw out what you think is more reasonable (2000? 2500?), and I can show how the amounts change.

@fractalvisions you’re 100% spot on, the best answer may be no shield at all and should be considered when evaluating optimal strategy.

KGB

The current 3-Star special offer is the first time they've offered something like this.

For $24.99 you get 48 covers (3 covers of 16 3* characters) + 2000 HP and 10K ISO. It's eerily reminiscent of the 2* offer that gives 3 covers of all 2*s ( 14 I believe including Bagman ).

As a comparison Logans Loonies gives 2900 HP for 19.99 so 2000 HP costs ~13.78. That means 48 covers is priced at ~11 dollars or 23 cents a cover and a fully covered 3* (13 covers) should be worth $2.99 (13x23 cents).

It's never been this cheap. That's more proof in my mind that 3s are the new 2 and that the 1* tier isn't long for this world.

KGB

Ares76

I only ever managed to place 1st once in a CL10 PvP slice. It was for Wasp shards but it felt great!!! How many shields did I burn? I don't know. Since I'm HP-positive that doesn't matter to me.

If you are constantly ranking Top 10 you will get 100 shards per 5*. That means you have to play through 5 rotations to get one cover. How long does that take nowadays? 2 years or more?

My main reasons for using shields in PvP are to get the 18 CP reward early (without having to grind to 50 wins) and to place good as individual and as alliance.

As mentioned before you could also be lucky and join very late into a bracket that has just flipped and get a top placement with no shield or a 3 hr shield. That would certainly be a net gain but I wouldn't take such a risk.

helix72

@KGB said:
The current 3-Star special offer is the first time they've offered something like this.

For $24.99 you get 48 covers (3 covers of 16 3* characters) + 2000 HP and 10K ISO. It's eerily reminiscent of the 2* offer that gives 3 covers of all 2*s ( 14 I believe including Bagman ).

As a comparison Logans Loonies gives 2900 HP for 19.99 so 2000 HP costs ~13.78. That means 48 covers is priced at ~11 dollars or 23 cents a cover and a fully covered 3* (13 covers) should be worth $2.99 (13x23 cents).

It's never been this cheap. That's more proof in my mind that 3s are the new 2 and that the 1* tier isn't long for this world.

KGB

This is another thing the conversion model allows one to do--evaluate monetary offers. It allows for the conversion of every deal into a number of 5* expected covers which you can look at on a per dollar basis to evaluate which deals, from a strictly economic standpoint, are worth buying. For example, the recent Eternal Youth deal, which for $34.99 USD gave you 5500 HP, 3 CP, 2 LTs, 525 Sersi shards, and 450 Kate Bishop shards was, relatively speaking, an awesome deal per 5* cover ($4.65). On the other end, the Immortal Fear deal which for $39.99 USD gave you no HP, 2 CP, 2 LTs, 525 iHulk shards and 450 Juggernaut shards was a relatively poor deal per 5* cover ($17.85). Of course, I think most would say iHulk is more important than Sersi but now you can decide if he's $13.20 per cover more important.

These amounts are based on the assumptions as stated in my original post.