1. Correction to Course Depth-Focus Formula

The previously-formula for course depth and focus given in my note of 5 August was wrong: The correct formula is:

If[Depth=1, Which[N<9,F=LD; N<16,equal prob. of LD & M; N<24, equal prob. of M & UD; UD];

If[Depth=2, Which[N<16, M; N<24, equal prob. of M & UD; UD];

If[Depth=3, UD],

2. Equation (3) of Td_3.3

Fred discovered a discrepancy between equation (3) of Td_3.3 and the associated Excel formulation in HE.GDB.init. The Excel formulation is correct. Td_3.3 has been corrected and will be reissued.

3. Separation of Libraries and Information Technology

We agreed in Hong Kong that, to make things easier for the Player, libraries and information technology should be separate line items on the interface and separate variables in the Stage 2 resource allocation procedure. Existing initialization procedures should be used to get the initial values of the two variables: i.e., the combined total will be calculated first and then the split will be made. From that point on the two variables will be handled separately.

4. Quality of Internet Connection

Quality of internet connection, one of the items shown on the game interface, is a surrogate for investment in IT. We agreed that to be consistent with other such items we will provide for four categories. The "Quality of internet connection" response function provides a 0-100 scale for the underlying continuous variable, which is an input to a number of other response functions. We should map the continuous variable to the discreet one using equal intervals: i.e., 0-25->lowest category,...76-100->highest category.

5. Interest Rates and Debt Capacity

Short-term interest will be credited or assessed monthly, based on the current applicable short-term real interest rate and the rate of inflation and the operating reserve balance at the end of the prior fiscal year. Positive balances will earn interest and borrowing from the bank will require interest payments at the bank borrowing rate, which is higher than the bank rate for deposits. Short rates and inflation may change each month. so the normal procedures for projecting to year-end apply. Bank borrowing reflects a negative operating reserve and will be shown as such on the balance sheet.

Interest at the bank deposit rate should be added to the plant reserve at the beginning of each year (before the facilities model is run), based on the balance at the end of the prior year. We didn’t discuss this in Hong Kong but it’s provided for in Td_2.4.

There is a small change to the procedure for determining long-term debt capacity. Cell E46 of HE.Faciltiies calls for the "Maximum percent of available funds" to be applied to the overall funds balance. However, the overall funds balance is not consistent with the definition of "available funds" as used by the bond rating agencies. We approximate the definition by the following formula:.

The long-term debt capacity formula from E46 is:

where "maxDebtLtPct" and "maxLtIntPct" are invariant game parameters. (The value currently specified are 50% and 8%, respectively.)

It was agreed in Hong Kong that the game will terminate if bank borrowing exceeds the short-term debt capacity. Hence we define:

where maxStDebtPct = 10% and maxStExpPct = 50%. (All such values should be reviewed during beta testing.)

Both the short and long-term interest rates depend on the balance sheet at the end of the prior year.

• The nominal long-term rate equals the base long-term rate times a multiplier calculated by the "Long-term interest rate" response function, which is based on the amount of long-term debt outstanding in relation to long-term debt capacity.
• The real short-term borrowing rate equals the base short-term rate times a multiplier calculated by the "Short-term interest rate" response function, which outstanding bank borrowing in relation to short-term debt capacity..

Both response functions will be defined shortly.

1. Special Treatment for Athletes

In Hong Kong we agreed on the following translation from the interface to the parameters for traditional undergraduate student segment 5 [athletes]:

• If "special admissions treatment for athletes" = "High" then set utility for segment 5 to equal 10; if = "medium" then split the difference between the 10 and the previously calculated utility for segment 5.
• If "special aid for athletes" = "High" then set the aid figure for segment 5 to equal maxAid; if = "medium" then split the difference between the maxAid and the previously calculated aid for segment 5.

Utility and aid for the other segments are unchanged.

1. Level of Competition in Athletics

The initialization values of athletics revenue and expenditure depend on the level of athletic competition ("NCAA level") chosen for the PGI. Player can change the level of competition anytime during the game, but so far the specifications are silent as to the consequences of such changes.

Any change in NCAA level will become effective at the beginning of the following year. Once made, the change cannot be revoked until the following year. In other words, at least one year must elapse between changes, and an announced change cannot be canceled.

The change produces new revenue and expenditure values, which are calculated as follows:

• Using the initialization formulas, calculate the "change ratios" of revenue and expenditure for the new NCAA level relative to the old one, ignoring random elements if any.
• Also calculate an "expenditure normalization ratio" equal to current athletics expenditures divided by the initialization expenditure value for the old NACAA level.
• Multiply the change ratios times the current levels of revenues and expenditures to get the new values. Working from current levels in this way takes account of inflation and player decisions since the start of the game.
• Substitute the new revenue and expenditure values in place of the prior-year values in the RA procedure as it will be applied to the new year (the one in which the NCAA change will first become effective).
• Calculate a new "athletics expenditure norm" by multiplying the initialization expenditure value for the new NCAA level by the expenditure normalization ratio. This takes account of inflation, but it also incorporates the effect of past Player decisions in the norm for the new level. I regard this as acceptable in the interest of simplicity.

The new values of revenue and expenditure will enter the financial statements. The new expenditure and normal expenditure values will affect the win-loss probabilities and various satisfaction functions.

1. Upper Limit on the Research Overhead Rate

Players should not be allowed to set an upper limit for the research overhead rate that exceeds the rate that would be allowed by the Federal government given the prior year’s budget. The formula for the maximum rate is:

(sum of all operating expenditures except departmental expense and sponsored research)/(sum of except departmental expense and sponsored research)

Departmental expense and sponsored research are the first two lines of the expenditure statement. Operating expenditures excludes the transfer to plant and debt service. This maximum rate applies to the variable limit that is normally set by the Player, not to the hard-wired global limit.

2. Global Upper and Lower Limits on RA Policy Variables

Clarity in the interface requires that we set global upper and lower limits on the sliders. The limits and their labels will be hard-wired into the game code. Figures for each budget policy variable are provided in the "Lower limit" and "Upper limit" columns of Table 1. The other columns will be defined later. The policy variable list conforms to decisions made at our August meeting.

The (*) on surplus (deficit) indicated that this variable is again expressed as a percentage of the operating budget rather than the dollar figure I tentatively specified in Hong Kong. Although dollars may be more natural to the player, it would be impossible to scale dollar figures to the range of institutional budgets to be used in the game.

3. Growthrates for Use in Financial Projections

We clarified the procedures for calculating the growthrates used in financial projections. The following description uses t for the current year, t–1 for the prior year, and t+1 for the subsequent year.

• Projecting year t at the end of year t–1: use the growthrates from the Stage 1 and Stage 2 RA optimization and the current (i.e., year-end t–1) values for inflation and the other exogenous growthrates.
• Projecting to year-end for year t during the year: the general procedure is to assume that current-month values will persist to the end of the year. Adding this projection to the year-to-date value produces the projection. Special procedures apply to some variables as discussed in our meetings.
• Projecting for year t+1 during year t: apply the optimization values for year t and current values of the exogenous growthrates to the year-end projections for year t.

1. Penalties for Breaking RA Promises and Commitments

While budget optimization occurs only at the end of August, Players can change the budget policy variables anytime during the year. The Player can take one of three actions:

• implement the change immediately: takes effect in the month when the decision was made and continuing through the following fiscal year (the variable will be locked during the ensuing optimization, effect in the current month rather that the following one will make little difference to the Player and simplifies programming);
• promise the change for the next budget year: resets the target and upgrades the priority to "3" (the maximum value);
• commit to consider the change for the next budget year: resets the target and changes the target to the maximum of "2" (the middle value) and the value before the change.

The optimization input screens should display an icon that signals whether any of the above are active and if so which one.

The Player can override the "promise" and "consider" settings any time prior to the final optimization, but this will incur a penalty due to stakeholder disappointment and/or loss of operating efficiency. The third column of Table 1 specified the variable(s) where the penalty will be inflicted ("Penalty variable"), the fourth column gives the direction of change in the variable that will produce a penalty ("dir" in the formula), and the last column the maximum amount of effect per year ("max"). The (*) on certain penalty variables indicates that, because of the way the variable is constructed, the effect occurs in the first month of the fiscal year only.

Penalties will be applied as multipliers to the indicated functions. The generic formula for the multipliers follows. Let X be the value of the penalty variable after optimization and X₀ be the promised or committed value.

where upLim and loLim are the variable’s global limits and respF[_] is an s-shaped function that converts the argument to a 0-1 variable as specified on the response function sheet.

1. Other Cost-Rise

The "Other cost-rise" variable in the Stage 1 optimization has been added as an input to the "Administrative performance" and "Staff morale" functions. Other cost-rise already affects faculty performance through the "Academic support ratio."