So. Much. I don't understand

[Captain Striker (Empo)]

While I'm a bit of a perfectionist, I really dont get ANY of the whole special gem combining. Even with the pictures (insert unlucky brian meme). So I'm stuck with doing the next best thing END GAME WITH HIGH WIZARD LVL:

I distilled from all the info that feeding gems with lowergrade gems still ups the specials
And that adding pure gems to multi-colour gems that are from the same grade, usually result in better specials for that multi-color gem.

So What I'm doing is this:
Step 1: Grade 1 Orange Gem + Grade 1 Red Gem = Grade 2 Orange-Red Gem
Step 2: Grade 1 Orange Gem + Grade 1 Black Game = Grade 2 Orange Black Gem
Step 3: Grade 2 Orange-Red Gem + Grade 2 Orange Black Gem = Grade 3 Orange-Black-Red Gem
Step 4: Duplicate Grade 3 Orange-Black-Red Gem so you have 2
Step 5: Create a Grade 3 Orange Gem and a Grade 3 Black Gem
Step 6: Grade 3 Orange Gem + Grade 3 Orange-Black-Red = Grade 4 Orange-Black-Red Gem
Step 7: Grade 3 Black Gem + Grade 3 Orange-Black-Red Gem = Grade 4 Orange-Black-Red Gem

And you repeat the whole adding of the Orange and the Black to the duped tri-colour gems according to their grade. So when the tri-color gem is Grade 30, you add Grade 30 Orange and Black to it and it's dupe.

Since this makes the gems also deal a good amount of damage, you should throw quite a good number of bombs on the waves. Currently I'm bombing 50 Grade 1 gems on the early waves in order for the mobs to survive the mana farm and get to the kill gem.


Might not be perfect, but it still works wonders for me. So I just wanted to share a decent alternative.

[dholland]

Originally posted by 12345ieee:

A 2048c multiplies the cost of a gem by 2048 (minus a single combine cost, which can be neglected).
If the resulting gem has a different value, you have wrong settings on wGC (delay too low, infobox shown).

No. This is wrong. The 2048 mana combine recipe involves combining lower grade gems with higher (as in (2m+m)) which as I mentioned, adds to the cost. The settings of the program are not relevant to that, it's just about how the game does combines.

We already saw with my spec example (sorry about the g7 thing) that combining lower grades with higher adds to the cost of the gem compared to not doing that.

[12345ieee]

You are right to be concerned, it's all but obvious that you gain from doing that.
That's why I wrote a program to do the comparisons for me.

Turns out you need both "cost inflating" and "grade inflating" in the right proportions (aka, you need specific recipes).

----------------

Are you kidding me? A 2048c gets 2048 copies of the base gem and combines them together.
By DEFINITION this multiplies the cost by 2048.

It may not up the grade by log2(2048) = 11, but that's another story.

[dholland]

I don't want to get into an argument. There is nothing about cost inflating in the guide or the program help. But at this point, I give up. We're not making progress.

[darthmohawk1]

Let me try explaining things in different words.

Originally posted by dholland:

Originally posted by 12345ieee:

A 2048c multiplies the cost of a gem by 2048 (minus a single combine cost, which can be neglected).
If the resulting gem has a different value, you have wrong settings on wGC (delay too low, infobox shown).

No. This is wrong. The 2048 mana combine recipe involves combining lower grade gems with higher (as in (2m+m)) which as I mentioned, adds to the cost. The settings of the program are not relevant to that, it's just about how the game does combines.

We already saw with my spec example (sorry about the g7 thing) that combining lower grades with higher adds to the cost of the gem compared to not doing that.

Let's call the base gem for this example a g12, because that's the highest-grade gem you can create with just one click . A g12 gem, with Fusion level 60, has a cost and value of 110,544.

The most basic combination that could be considered a "2048" would be to "U" upgrade the base gem 11 times, which would increase the base gem from g12 to g23. This gem will cost 2,048 times as much, plus the cost of the 2047 combinations: 226,492,368.

Now let's compare that with the Mana Combine 2048 given in wGemCombiner. This process involves 185 separate steps of duplicating and combining, which eventually results in a gem 9 grades higher than the starting gem: g12 -> g21. This g21 gem, however, was made using the same materials as the g23 in the previous paragraph, just combined in a different order. Both gems have 2,048 instances of the base g12, plus 2,047 combines to glue them all together. (You can confirm this, if you have the patience, by adding the count of all the "m"s in the written recipe, and see that they add up to 2048.) Therefore, the g21 made with wGemCombiner will have the same cost as the g23: 226,492,368.

The correct comparison, therefore, is between the g21 gem and the g23 gem, as even though they have different grades, they have the same mana cost. I have a demonstratory image[i.imgur.com] which I believe should help in this regard. You may notice that the g21 gem created with wGemCombiner, although it is two grades below the g23 created with "U", has a significantly higher mana leech component.

[dholland]

I'm feeling so dumb, but happy! Some of what you've been telling me has finally sunk in. I now understand that you have to compare gems at equal mana value, and the grade actually doesn't matter. Although you could have a "cheap" gem with low cost per grade, as in equal grade combinations, and an "expensive" GC gem with high cost per grade, once you've upgraded/combined them to reach the same cost (whatever you can afford to spend), you just compare the gem power and that's it (assuming you don't make my mistake of comparing a gem below grade 7 with one of higher grade). It turns out true that the GC specs and combines are just so good (presumably with a computer search of all recipes up to a certain number of base gems), that the "free" combines of "cheap" gems just can't catch up the power of the GC gems. It's helped me to use the new version of GC, I've not had any failed recipes even at the default delay. Whereas before I didn't feel comfortable playing with mana lock, and yesterday got 179 billion XP in an Endurance, today I played with mana lock and got 583 billion XP gain (over 2.3 billion previously). I forgot to screenshot the XP screen but my wizard level bumped to 13,283 from 12k:

https://www.dropbox.com/s/ihfhgawkdut4w8x/Screenshot%202016-05-26%2021.04.54.png?dl=0

I used some quite high combines as you have a huge mana gain when the mana lock opens, so you have a lot to spend. It probably takes less time than constantly upgrading/combining without mana lock, and using the GC combines you can be quite efficient about it (although I didn't go mad and spend more than a couple of minutes on any recipe). I also used freeze and WoE to build up a high number of monsters going over each managem trap, and so more to be chain hit within the trap.

One thing I worked out in GC. With specs, the gem combine cost (48 with everything maxed) still greatly exceeds the g1 gem cost (6). So I worked out that I have to multiply the number of base gems by 54 (replace that value with the sum of the g1 cost and the cost of combining two gems for people without everything maxed) to estimate the final cost of the spec. With combines, usually the starting gem is much higher than g1 and you can mostly just neglect the combination cost because the base gem costs so much more, and you just have to multiply the base gem cost by the number of base gems in the combine to get the final cost. No doubt obvious to most people, but I'd got confused by some failed combines or varying cost specs, with the older version, which was probably because I used the default delay and this was not long enough to avoid errors.

[Captain Striker (Empo)]

Originally posted by dholland:

I'm feeling so dumb, but happy! Some of what you've been telling me has finally sunk in. I now understand that you have to compare gems at equal mana value, and the grade actually doesn't matter. Although you could have a "cheap" gem with low cost per grade, as in equal grade combinations, and an "expensive" GC gem with high cost per grade, once you've upgraded/combined them to reach the same cost (whatever you can afford to spend), you just compare the gem power and that's it (assuming you don't make my mistake of comparing a gem below grade 7 with one of higher grade). It turns out true that the GC specs and combines are just so good (presumably with a computer search of all recipes up to a certain number of base gems), that the "free" combines of "cheap" gems just can't catch up the power of the GC gems. It's helped me to use the new version of GC, I've not had any failed recipes even at the default delay. Whereas before I didn't feel comfortable playing with mana lock, and yesterday got 179 billion XP in an Endurance, today I played with mana lock and got 583 billion XP gain (over 2.3 billion previously). I forgot to screenshot the XP screen but my wizard level bumped to 13,283 from 12k:

https://www.dropbox.com/s/ihfhgawkdut4w8x/Screenshot%202016-05-26%2021.04.54.png?dl=0

I used some quite high combines as you have a huge mana gain when the mana lock opens, so you have a lot to spend. It probably takes less time than constantly upgrading/combining without mana lock, and using the GC combines you can be quite efficient about it (although I didn't go mad and spend more than a couple of minutes on any recipe). I also used freeze and WoE to build up a high number of monsters going over each managem trap, and so more to be chain hit within the trap.

One thing I worked out in GC. With specs, the gem combine cost (48 with everything maxed) still greatly exceeds the g1 gem cost (6). So I worked out that I have to multiply the number of base gems by 54 (replace that value with the sum of the g1 cost and the cost of combining two gems for people without everything maxed) to estimate the final cost of the spec. With combines, usually the starting gem is much higher than g1 and you can mostly just neglect the combination cost because the base gem costs so much more, and you just have to multiply the base gem cost by the number of base gems in the combine to get the final cost. No doubt obvious to most people, but I'd got confused by some failed combines or varying cost specs, with the older version, which was probably because I used the default delay and this was not long enough to avoid errors.

There are recipies to be found online? That's great to read. Any advice on where to go? Because I only just discovered that my 'old school' way is sub-optimal and I'd like to be squeezing out more exp.

[darthmohawk1]

Originally posted by Captain Striker:

There are recipies to be found online? That's great to read. Any advice on where to go? Because I only just discovered that my 'old school' way is sub-optimal and I'd like to be squeezing out more exp.

The most basic ones (IE the only ones you can do by hand) are in Bilbo's extreme guide. For anything higher than 64c, if you download wGemCombiner, the program includes a dropdown menu with combines going to up to like 3 million or so.

[BilboCGL]

Originally posted by Captain Striker:

There are recipies to be found online? That's great to read. Any advice on where to go?

Absolutely no idea...[/sarcasm]
That's a FAQ...

[Captain Striker (Empo)]

Originally posted by BilboCGL:

Originally posted by Captain Striker:

There are recipies to be found online? That's great to read. Any advice on where to go?

Absolutely no idea...[/sarcasm]
That's a FAQ...

I am truly amazed that this game still has activity after coming out such a long time ago. Must be the kind and helpful atmosphere of the people playing who are able to interpret everything others write correctly. :) [/amidoinitrite?]