Most of the Haas ICE is bioroid, which means they are run by artificial intelligences. The effect of this is that the runner can usually break subroutines on the ICE by spending clicks rather than using a breaker. They can work around them, or talk their way past them.

This is a bit of a weakness, so I need to make sure I also have some ICE with requires the runner to spend credits.

The way to think about ICE is in terms of two sets of properties:

1. Binary vs Analog
2. End-The-Run vs Taxing

Every piece of ICE fits into one of the four categories made up by these properties.

‘Binary’ means the ICE can be completely bypassed once the runner has the correct breaker. ‘Analog’ means that it will still do something even if its subroutines are all broken.

‘End-The-Run’ (ETR) means just what it says, one of the subroutines can end the run. ‘Taxing’ means that it can’t end the run, but the runner must pay some kind of extra cost to bypass it.

If you take these four properties, you get the following four categories:

Binary ETR: These are usually cheap and good for putting on R&D or HQ early on, for example Enigma. They are rarely useful later on, as they get turned off once the runner has the matching breaker.

Binary Taxing: These have no ETR subroutines, and tend to be cheap to pass with the right breaker, e.g. Viktor 1.0.

Analog ETR: These are usually expensive to break, e.g. Tollbooth or Wall of Thorns, and are good for central servers as they make runs expensive for the 20% chance of hitting an agenda in HQ for example.

Analog Taxing: E.g. Heimdall, where the runner usually pays the tax (clicks) rather than the credits to break.

The bioroid ICE is mostly binary. The more expensive ICE can be considered taxing in the sense that they have a high strength and the runner must pay to boost her breakers’ strength, but the bioroid ‘click’ weakness kind of undoes this.

I think I need to add in some taxing ICE, which will make it expensive for the runner to try breaking into my servers.

Binary ICE is usually cheap and is good for early on, before the runner has their full rig ready.

In general, as a corp you never have ICE that can completely keep the runner out of a server. They can always get in eventually. They just need to wait to install the correct breakers and then save up the credits.

The trick is to make it expensive for the runner to access my servers, so they can’t run so often. Then I will have a window of a turn or two to quickly score an agenda. Early on, I can use cheap binary ICE to keep the runner out while they don’t have their full rig yet.

This kind of deck is known as a ‘glacier’ deck, where you slowly build up big ICE.

The recommendation is for 17-20 ICE for a deck of 45-49 cards. So if I split into code gates, barriers and sentries and have 5 or 6 of each, and have some cheap binary and expensive taxing ICE in each, then I can have the following:

Code Gate: Binary = Enigma, Datapike. Analog = Viktor 1.0, Viktor 2.0.

Barrier: Binary = Wall of Static, Bastion. Analog = Heimdall 1.0, Heimdall 2.0.

Sentry: Binary = Rototurret. Analog = Ichi 1.0, Ichi 2.0.

The actual recommended figure is around 40% ICE in your deck. Here is a table of the percentages with different amounts of ICE in different seizes of decks.

	45 Cards	46 Cards	47 Cards	48 Cards	49 Cards
17 ICE	37	37	36	35	35
18 ICE	40	39	38	38	37
19 ICE	42	41	40	40	39
20 ICE	44	43	43	42	41
21 ICE	47	46	45	44	43

The general principle is that you should always have a 49 card deck as the corp. This is because it makes the density of your agendas as low as possible, making it hard for the runner to hit them in runs on R&D and HQ.

So this would mean I should have 20 ICE (to ensure I have 40% ICE in my deck) and I can add in some additional non-ICE cards, such as Alix Talbot for more funds.

Now to see how it plays.